Which divisions moved into Section J?

Divisions moved into new Section J in ISIC Rev. 5, UN Statistics Division, as part of the split of old Section J, and the authoritative list is the UNSD Rev. 5 structure itself. This guide does not list division numbers from memory, because a wrong number here would misdirect a reader’s code work. The safe method is to open the UNSD structure, filter to Section J, and read the divisions and their explanatory notes there, checking each activity in words before using any number.

Why is there no memorised list in this guide?

There is no memorised list here by design, and the reason is practical rather than cautious for its own sake. A division number copied from memory, or from a secondary summary, can be wrong in a way that looks authoritative: the number is correctly typed, the section letter is right, and the sentence around it reads smoothly, yet the division it names may cover different activities from the ones the reader described. That kind of error is hard to spot in proofreading and easy to propagate, because the next reader copies the same number on the strength of the first guide. The authoritative list lives in the UNSD Rev. 5 structure, and it can be read there, filtered to Section J, with each division and its explanatory note in front of the reader. A guide that points to that list, and explains how to read it, serves the reader better than a guide that prints a list the reader cannot verify without opening the same structure anyway.

How do you read the Section J divisions in the UNSD structure?

Open the structure and filter to Section J, then work through its divisions one at a time, reading each division in words before looking at its number. For each division, ask what kind of content work it covers, read its explanatory note, and check the groups and classes beneath it, because the division title alone is a summary and the finer levels carry the boundaries. Where two divisions look close, read their excludes lists against each other: the lists will often show, plainly, which kind of content activity belongs in one division and which belongs in the other. Keep a note of the activity in words beside each division considered, so the reasoning survives beyond the moment of reading. If a division cannot be justified from the evidence available, do not force it. Report at the section level, or at the broader level that is clearly correct, and record that the finer detail was not available. That record is part of the result, not an admission of failure.

What happens to a comparison when a division list is copied from a summary?

A copied list creates two problems at once. The first is accuracy: a summary may reflect an earlier revision, a national variant, or a simplification that drops the very distinctions the reader needs, and none of those problems is visible in the copied numbers themselves. The second is comparability across tables: if one table uses a copied list and another uses the UNSD structure directly, the two tables may report different division totals under the same labels, and a reader comparing them will see a difference that reflects the source of the list rather than any difference in content work. The damage spreads because Section J figures feed wider tables: national accounts, business registers, and related product and occupation tables all inherit the division totals they are given. The remedy is to derive the list, every time, from the same source, with the version label, ISIC Rev. 5, UN Statistics Division, attached, so that any two readers who follow the same steps reach the same divisions.

What does a sound Section J division note contain?

A sound note starts with the activity in words, names the divisions considered in the UNSD structure, states the division or broader level finally used, and gives the reason in terms of the explanatory note that supported it. It also names the version, the level, the reference period, and the source, and it flags any break in series where figures from different revisions appear together. A survey team applies the same discipline upstream: questionnaire, coding manual, and published table must all point to the same definitions, or results drift between collection and publication even when every individual coder worked carefully. A reader who writes that note at the time of use saves the next reader from reconstructing decisions that were never recorded, and that saved reconstruction is often the difference between a table that can be reused and a table that must be rebuilt.

This guide stays within statistical classification only, and it settles nothing for an individual reader: it does not determine any reader’s code, principal activity, or registration, and any code for a specific record is to be confirmed with the national statistical institute or business register that holds the record.