How to tell whether recovery is working

Reconstruction reporting gravitates towards the easiest numbers: money pledged, money disbursed, projects approved. Those are input measures. They tell you about the administration of a programme, not about whether anything got better.

What an outcome measure looks like

Housing units restored to habitable standard, and occupied — the second half matters, because a repaired empty flat is not housing.

Households connected to water, heat and electricity, measured at the connection rather than at the plant.

Schools open with pupils in them, and hospitals performing procedures.

Hectares of land released after demining and returned to production.

And population present in a district, which is the summary indicator all the others feed into.

Why input measures dominate anyway

Because they are available quickly, they are unambiguous, and they make a programme look busy. An outcome measure lags, moves slowly and sometimes moves the wrong way for reasons outside the programme.

That is precisely why an outcome measure is worth publishing: it is harder to manage and therefore more informative.

The publication discipline

A fixed set of indicators, a fixed schedule, the same definitions each time, and revisions marked as revisions. Changing the indicator set between reports is how a programme conceals a problem without anybody lying.

The test

If a report cannot say how many people are living in restored housing, it is a report about spending, not about recovery.

I judge whether a project is really moving by the work delivered rather than by the money spent; on the equipment side that is the most basic measure there is. Input measures are easy to report and say nothing: what matters is not how many machines arrived but how many are running. The same holds for reconstruction — not the number of contracts but the number of schools open.