ICTlogy

Impact Assessment
Tim Unwin.

We need to know who is driving impact assessment and what for.

• Impact: what has changed. Long term change.
• Output: what we did. Goods or services produced.
• Outcome: what happened. Changes in the use of those outputs.

The Logic Model: inputs are put to work in activities, generating outputs and, at the end, outcomes. The log frame approach crosses goals, purposes, outputs and outcomes and analyzes them during the different processes, with different indicators, etc.

Impact would include the long-term intended and unintended consequences of our action, of the outcomes.

The problem with some impact assessment models is that they might measure the performance in managerial terms, but not much in developmental terms. in fact, impact assessment might end up in programme evaluation, not in impact evaluation itself, thus making ends often driving the means. On the other hand, criteria might be set in advance by the evaluation promoter.

Terry Smutylo’s Output Outcome Downstream Impact Blues.

But there are reasons for focus on impact, like where and when to scale a pilot project.

Cause and effect are very difficult to establish, especially attribution, that is, identifying what is the cause. If the impact is placed in the (very) long-term, things get even more difficult. Indeed, experimenting in the field of ICT4D can neither be possible nor be ethically accepted (i.e. “are you going to experiment with people’s welfare?”), making research and measuring the impact of that research really challenging.

In development, quite often it is more about the “why” rather than the “what” happened. This is especially relevant when different contexts might bring different impacts for the same cause given.

If nevertheless impact assessment is to be done, we need to have a clear benchmark of what has been done, test with different samples and contexts, etc.

Previous post: Horizon Report: Ibero-American Edition 2010

Next post: Fifth Annual ICT4D Postgraduate Symposium (II). Gender