Why the “Average Price per Hectare” Can Be Misleading

Calculate the price per hectare of agricultural land “naively” — a plain arithmetic mean across all registered purchase-sale deals — and you get UAH 168,633/ha. The real all-time weighted average price across Ukraine on LandData Explorer is UAH 43,700/ha. That’s a 3.9x difference. We break down why a simple average is this misleading, and what to do about it.

The reason lies in the shape of the dataset itself. Let’s look at how prices are actually distributed across 457,733 purchase-sale deals over the entire observed period (July 2021 – August 2026):

• under UAH 50K/ha288,333 (63.0%)
• UAH 50–200K/ha124,753 (27.3%)
• UAH 200–500K/ha26,779 (5.9%)
• UAH 500K–1M/ha10,200 (2.2%)
• UAH 1–50M/ha7,613 (1.7%)

Data covers all registered agricultural land purchase-sale deals with a known price and area, July 2021 – August 2026. Source: State Geocadastre land relations monitoring.

The bulk of deals is entirely expected: 63.0% sold for under UAH 50K/ha, another 27.3% in the UAH 50–200K range. But beyond that, the distribution stretches into a long tail: 1.7% of deals fall in the UAH 1–50M/ha range, and there are a further 55 deals (0.01%) priced above UAH 50M/ha. The most extreme record in the dataset is telling: a 0.0081-hectare plot sold for UAH 23 million — which works out to UAH 2.84 billion per hectare. This is most likely a registration error (a mixed-up area or price unit) rather than a genuine market deal, but formally it enters the dataset the same as any other transaction.

A plain arithmetic mean weighs every deal equally, regardless of whether it’s typical or not. So each such “outlier” pulls the result up out of all proportion to its real significance for the market: a hundred normal deals at UAH 40K/ha won’t outweigh one record at UAH 2.84 billion. That’s exactly why a naive average across the whole dataset gives UAH 168,633/ha — this isn’t a calculation error, but the predictable consequence of a method that isn’t built for such a skewed distribution.

Different statistical methods handle this problem differently. Let’s compare four approaches applied to the same 457,733 deals:

The first three are calculated directly from the full dataset, with no filtering of anomalies. The fourth is the current all-time weighted average price across Ukraine on LandData Explorer.

The median sits at the opposite extreme: UAH 38,755/ha, 11% below the real price. That’s because the median treats every deal as equally significant regardless of plot size — and small plots include a higher share of low-priced deals, so the median systematically understates what a typical hectare is actually worth on the market.

Interestingly, even a plain weighted average (the sum of all prices divided by the sum of all areas, with no filtering at all) gives UAH 53,089/ha — already much closer to reality than the naive average, but still 21% above the actual site price. That’s exactly why weighting alone isn’t enough: it also takes consistent filtering of anomalous records before the result is calculated.

For anyone reading news or reports about the “average price of land,” this is a good reason to check exactly how that number was calculated. A simple average over an unfiltered set of deals can overstate the real price several times over — and this isn’t a rare, exceptional case but a systematic feature of land market statistics, also confirmed by academic research into the same phenomenon using 2021–2023 data: over 20% of deals there turned out to be statistical outliers, and a naive average produced a result 2–3 times higher than the real one. On LandData Explorer, the price per hectare is calculated not as a simple average but with a weighted method resistant to exactly this kind of anomaly.