How to Evaluate a University Partnership: Criteria, Indicators and Common Traps

Partnerships · By the Open Quad editorial team · Published 2026-10-19

Four panellists seated at a conference table with microphones answering questions — Panellists answer questions at a university partnership forum in Jakarta. Photo: USAID Indonesia, public domain (US government work).

University partnerships are easy to describe and hard to judge. A report can list workshops held, staff trained and agreements signed without saying whether anything changed. Whether you are a university leader deciding to renew a partnership, a funder, or a researcher studying the field, these are the questions and tools that separate activity from results.

The six criteria

The OECD Development Assistance Committee (DAC) published its evaluation criteria in 1991 and revised them in 2019, adding "coherence" to the original five. They are used, in some form, by most bilateral donors, the European Commission and the World Bank, and they translate well to university partnerships.

Criterion	Question for a university partnership
Relevance	Does it address a priority of the partner institution and its country, not only of the funder or the foreign university?
Coherence	Does it fit with the partner's own plans and with other programmes working on the same problem?
Effectiveness	Did it achieve its stated objectives, such as a new degree programme running or a number of staff completing doctorates?
Efficiency	Were results achieved at a reasonable cost and in reasonable time, compared with alternatives?
Impact	What wider, longer-term changes followed, intended or not, for graduates, employers, communities or policy?
Sustainability	Will the benefits continue when outside funding ends?

Write the results chain first

Most evaluation problems start with a vague objective. Before choosing indicators, set out the logic of the partnership in four steps, sometimes called a logic model or theory of change:

Inputs and activities. Money, staff time, visits, joint courses, equipment, scholarships.
Outputs. What the activities directly produce: a revised curriculum approved by the senate, twelve lecturers enrolled in doctoral programmes, a laboratory equipped.
Outcomes. The changes those outputs should lead to at the institution: graduates with skills employers recognise, more research grants won by partner staff, a programme accredited.
Impact. The broader changes the funder ultimately cares about: better farm advice, more qualified nurses, policy based on local evidence.

Each arrow in the chain rests on an assumption, for example that trained lecturers will stay at the university, or that employers will hire graduates of a new programme. Good evaluations name these assumptions and test them.

Indicators that mean something

Pick a small number of indicators at the output and outcome levels, define them precisely, and record a baseline at the start. Useful examples, grouped by the university's three missions:

Teaching: share of academic staff holding a doctorate; number of programmes revised and formally approved; enrolment and completion rates in those programmes; graduate employment surveys; external accreditation results. (For business programmes, see how partnerships align business education with employers' needs.)
Research: peer-reviewed publications with a partner-institution author as lead or corresponding author (not just as one name among many); competitive grants won by partner staff; doctoral students supervised locally.
Institution and service: policies adopted (for example on quality assurance or research ethics); budget lines the institution now funds itself; extension or clinical services delivered; partnerships with employers or government.

Activity counts, such as workshops held or visits made, are useful for management but are not results. A report that contains only those should prompt more questions.

Where the data come from

Partner institutions' own records (staff files, enrolment and graduation data, budgets) are the most important source and should be collected with their agreement. Bibliographic databases show publication patterns. Surveys and interviews with staff, students, graduates and employers capture what records cannot. National data, such as those published by the UNESCO Institute for Statistics, help put institutional figures in context. For US-funded projects, many past evaluations are archived in USAID's Development Experience Clearinghouse and can be used to compare similar programmes.

Common traps

What makes evaluations misleading

No baseline. Without a starting point, "forty staff now hold doctorates" says nothing about change.
Attribution by default. A rise in publications may come from a national policy or another donor. Ask what would have happened anyway.
One-sided evidence. Reports written only by the foreign university, or only by the funder, miss the partner's view.
Measuring too early. Doctorates take four or more years; employment effects take longer. An end-of-project evaluation cannot see impact.
Ignoring what stopped. Look back two or three years after funding ends to see what survived.

Mutual benefit is a criterion too

Partnerships are rarely one-way. Foreign universities gain research sites, data, teaching material, international experience for students and staff, and sometimes reputation. Being explicit about what each side gains makes a partnership more honest and more durable, and it is increasingly expected by funders and by the partner institutions themselves. Principles published by research councils and university associations stress equitable authorship, shared ownership of data and fair sharing of budgets.

A short checklist

Is there a written results chain, agreed by both partners?
Are there five to ten defined indicators, with a baseline?
Who collects the data, and does the partner institution keep it?
Is there an independent evaluation, and is it published? (National quality assurance reviews can provide useful external evidence.)
What is the plan for each cost after the grant ends?

For background on the programmes these questions were developed for, see what university partnerships for development built.

Sources

OECD, DAC evaluation criteria (1991, revised 2019).
USAID Development Experience Clearinghouse: dec.usaid.gov.
UNESCO Institute for Statistics: uis.unesco.org.
World Bank, tertiary education: worldbank.org/en/topic/tertiaryeducation.