Why Demographic Data Need Clear Labels

Two numbers can look identical on a page and mean entirely different things. Population: 47.8 million might be a completed census count, a statistical estimate for a year between censuses, a projection for a future date under stated assumptions, or a modelled figure produced by whoever published the page. Printed in the same typeface, they are indistinguishable.

Almost every serious error in public demographic argument begins there. This is the standard this publication applies, and the reasoning behind it.

The four statuses

OFFICIAL. A figure produced by a national statistical authority or an equivalent official body, describing something that has already been measured. A completed census count. Registered births and deaths for a closed year. A published labour force survey result.

Even here, precision has limits. Censuses undercount some populations more than others - typically the mobile, the recently arrived, the poorly housed and those with reason to avoid officials. Register data and census data can disagree. Definitions change between rounds. OFFICIAL means the figure carries official authority and describes the past, not that it is exact.

ESTIMATED. A figure for a period not directly measured, produced by an accepted method - typically a base population carried forward using recorded births, deaths and migration. Inter-censal population estimates are the ordinary case. These are usually reliable and routinely revised when the next census arrives. An estimate revised downward is a system working correctly, not a scandal.

PROJECTED. A figure for a future date, produced by applying assumptions to a current population. It is conditional by construction. The correct reading of a projection is never the population will be X but on these assumptions, the population would be X. Serious producers publish variants for exactly this reason: the UN's low, medium and high fertility variants are not hedging, they are the honest expression of the range.

NATION CONTINUITY MODEL. A figure this publication has calculated itself, for a scenario or a horizon that official producers do not cover. It carries the lowest evidentiary weight of the four, it is always labelled, its assumptions are always stated, and it is never presented as the output of an official body.

Where the errors come from

Five recurring failures, in rough order of how much damage they do.

Status laundering. A projection is quoted in a report; the report is quoted in an article; the article is quoted in a debate. At each step the qualification erodes. By the fourth repetition, a conditional statement about 2075 has become a fact about the future. The 2100 figures in UN World Population Prospects are among the most frequently laundered numbers in public discussion.

ADVERTISEMENT

Vintage collapse. Figures from different reference years are compared as though contemporary. A 2011 census result set against a 2023 estimate produces a change that is partly real and partly an artefact of the twelve years between them. Every figure needs its year attached, always.

Definitional mismatch. Two countries publish urban population and mean different things, because the threshold defining a settlement as urban is set nationally. The same applies to household, migrant, unemployment and - as Brazil demonstrates at length - ethnicity. Numbers that share a label do not necessarily share a definition.

Precision inflation. A projection reported to the nearest thousand implies a confidence the method cannot support. Rounding is not sloppiness; it is an accurate representation of what is known.

Denominator drift. A percentage changes when either number moves, and reporting frequently attributes the change entirely to the numerator. A group's share of a population can rise because that group grew or because everyone else shrank, and those are different findings with different causes. Share and count should always be reported together.

The horizon rule

One rule follows directly and this publication treats it as absolute.

The United Nations World Population Prospects 2024 projects to 2100. It does not project beyond 2100. Therefore no figure for any date after 2100 can be described as a UN projection, and any publication presenting one as such has made an error or is misrepresenting its source.

Where this magazine publishes a figure for 2126 - a horizon chosen because it sits a century beyond the magazine's founding - that figure is a NATION CONTINUITY MODEL. It is labelled as such, it carries its assumptions, it is shown with a range rather than a single value, and it is presented in visual styles distinct from those used for official and projected data so the distinction survives being screenshotted.

That last clause is not decorative. Charts travel further than the articles containing them, and a chart that relies on surrounding text to explain its status will eventually be seen without it. Labelling has to be inside the image.

The source hierarchy

Where sources conflict, this publication prefers them in this order: the national statistical authority of the country concerned; the relevant ministry or official register; the established international compilations - UN, Eurostat, OECD, World Bank - with their reference year stated; peer-reviewed research; reputable secondary reporting.

News coverage of a statistic is not a source for that statistic. It is a pointer to one. Where a primary source exists, this publication cites the primary source.

International compilations deserve a specific note. They are indispensable for comparison because they harmonise national figures onto a common definition - and that harmonisation means the figure they publish for a country may differ from the figure that country's own statistical office publishes. Neither is wrong. They are answering slightly different questions, and a responsible citation says which one is being used.

Why this belongs in a culture magazine

Because demographic numbers are among the most politically active statistics in circulation, and because the mechanism by which they become misleading is almost never fabrication. It is the quiet removal of a qualifier.

A magazine arguing that cultural continuity is a serious subject has a particular obligation here. The demographic case is the one most readily borrowed for arguments this publication does not make, and the borrowing works precisely by stripping the labels. Keeping them attached is not pedantry. It is the difference between reporting and supplying ammunition.