Reading Ukrainian statistics properly

I use Ukrainian statistics constantly and I have developed a set of habits for reading them, which I set out here because most misreadings I see are avoidable.

Where the data is good

Trade statistics, because customs records every consignment. Monetary and banking data, because the central bank collects it directly from supervised institutions. Budget execution, because it is published transactionally. Procurement, because it is published in full.

These are administrative data — collected as a by-product of a process rather than by asking people — and administrative data is generally the most reliable kind.

Where it is thin

Anything touching the informal economy: employment, small business turnover, household income. A substantial share of activity is not recorded, so the level is understated and the trend is more reliable than the level.

Regional data, which depends on where a company is registered rather than where it operates.

And, since 2022, anything requiring a survey of the population, because the population moved.

The habits I use

Trust changes more than levels. A series that is consistently mismeasured still shows direction correctly.

Cross-check with a physical indicator: electricity consumption, rail freight tonnage, cement dispatches, customs volumes. Physical measures are hard to misreport and they discipline a financial series.

Read the methodology note. It is short, it is published, and it usually names the limitation the headline hides.

The general rule

A number without a definition is not information. Two people disputing Ukrainian economic data are usually using two different definitions and neither has checked.

I use the same habits: look at the series rather than a single figure, follow the revisions, and where possible verify the same quantity from two different sources. Where the data is thin is known — services, small business, informal employment. The rule is that you trust not the number but how it was collected.