The country cost figure is off by a median of 16.5%.
Including ours.

What the gap costs a reader

A gap of 16.5% on a $2,000 monthly budget is about $330 a month, or roughly $4,000 a year — enough to move a destination from affordable to not, and enough to sink a plan that was built on the headline. The direction is not consistent either, which is what makes it dangerous: 33 of the 62 cities come in under their country figure and 24 come in over it. A reader cannot apply a safety margin in one direction and be done.

Why it happens: there is no single price in a country

The obvious explanation — someone did the arithmetic badly — is not what we found. The second measurement is the one that explains the first. Inside a single country, the most expensive city we track costs a median of 43% more than the cheapest one. In UAE (Dubai) it is 121%. No single figure can represent a range like that, so every country headline in every ranking is describing a place that does not exist: the average of somewhere expensive and somewhere cheap, which is typically neither.

The full measurement, all 62 cities

Sorted by gap, cheapest relative to its country headline first. This is the whole finding — there is no subset we left out, and the same numbers ship as a CSV column named gap_vs_country_pct so you can recompute rather than trust us.

What this is not

This is two layers of our own data disagreeing with each other, not us catching anybody out. Both numbers are the midpoint of a published range: the country figure is a typical monthly budget for a retired couple living moderately, and the city figure is the same idea measured in one place. A midpoint is not a census. What is measurable here is the size of the disagreement, and 41% at the extreme is far too large to be rounding.

Two things follow, and they are the only claims we will make. First, a country-level cost figure should be read as a bucket label, never as a budget — including on our own country pages, which is why every one of our 62 city pages now prints its own gap against the parent. Second, any ranking that orders countries by cost is ordering buckets; the ordering can be right while the number attached to it is wrong for the place you would actually live.

The city set is what we track, not a random sample: 62 places across 17 destinations, chosen because retirees actually go there. A different set would give a different median. The direction of the finding is unlikely to change — a median gap this size does not come from city selection — but we would rather you check that yourself than take it from us.