How do you read a four digit code?

A four digit NACE code is read by splitting it into its division, group and class parts, and then finding its section in the structure table. The Eurostat manual gives the model example: code 20.51 for manufacture of liquid biofuels, where 20 is the division, 20.5 is the group, and 20.51 is the class. The section letter, section C in that example, does not appear in the code itself. Reading a code therefore means reading the numeric path, finding the section that contains the division, and reading the explanatory note for the class at the end of the path.

What are the steps for splitting a code?

The first step is to find the dot. The dot sits after the second digit. The two digits before the dot are the division code. The digit after the dot, together with the two division digits, forms the group code. The final digit, added to the group code, forms the class code. In code 20.51, the division is 20, the group is 20.5, and the class is 20.51. In code 75.00, the division is 75, the group is 75.0, and the class is 75.00. The zeros in the second example are meaningful: they show that the division is not split, and that the group and class carry the same heading as the division.

The second step is to find the section. The structure table in the manual lists sections with their divisions, groups and classes in order. Find the division number in that table, and read the section heading above it. That section letter is the section for the code, even though it is not written in the code. A dataset that shows a section letter beside a code has added the letter for presentation, from the same structure table. A dataset that shows only the numeric code requires the reader to do the lookup, and a reader must not guess the section from the division number alone, because division numbers repeat the pattern across sections only in the sense that each division sits in one section, and the section must be looked up.

Part of code 20.51	Level	How to find it
20	Division	Two digits before the dot
20.5	Group	Division code, dot, third digit
20.51	Class	Group code, fourth digit
C	Section	Look up division 20 in the structure table

What must be checked beside the code form?

The code form must be checked against the version. A code that is valid in NACE Rev. 2 may have a different meaning, a different section letter, or no equivalent in NACE Rev. 2.1, and a code that is valid in NACE Rev. 2.1 may not have existed in NACE Rev. 2. The correspondence tables record the relationship between versions, and the version label on a dataset or a table is part of the code reading. A code without a version label is incomplete, because the same digits can sit in different structures in different versions. The manual table comparing the two versions shows the scale of the change: 22 sections against 21, 87 divisions against 88, 287 groups against 272, and 651 classes against 615. Those differences mean that a code must always be read with its version stated.

The code form must also be checked against national extensions. A national version usually adds a fifth digit for national purposes, according to the manual. A five digit code is therefore a national code, not a European class code, and the fifth digit must be checked in the national publication. A reader who treats a five digit national code as a four digit European class, by dropping the fifth digit without checking the national note, may lose the national detail that the fifth digit carries, and may misread the scope of the entry.

How is the meaning of a code found?

The meaning of a code is found in its explanatory note, not in its digits. The digits show the place in the hierarchy. The note states what the category includes, what it also includes, and what it excludes, with cross references to other classes. The manual states that the entries in the NACE index are an integral part of the notes, and that the index lists detailed activities. A reader who needs the meaning of a code should therefore read the note for the class, read the notes for the group and division above it where the boundary is unclear, and check the index entry where one exists. The heading alone is a label. The note is the definition.

A worked reading, using only the manual example, shows the sequence. Code 20.51 splits into division 20, group 20.5, and class 20.51. Division 20 sits in section C in the structure table. The heading for class 20.51 is manufacture of liquid biofuels. The note for that class, in the explanatory notes part of the manual, states the inclusions and exclusions that define the class. A reader who stops at the heading has read the label. A reader who reads the note has read the definition, and is in a position to compare it with the correspondence table where a move from the earlier version is in question.

What does reading a code not do?

Reading a code does not assign the code to a unit, and it does not confirm an entry. The manual states that a unit is assigned one code in a statistical business register according to its principal activity, after considering all its activities. That assignment uses the notes and the classification rules, and it is made by the institute or register that holds the record. A reader who can split a code and read its note has learned how the classification describes an activity. That reader must still confirm an entry with the institute or register, and must not present a code read from a table as the code for a specific business.

Split the code at the dot, look up the division to find the section, read the class note for the meaning, and confirm any entry with the national statistical institute or the business register that holds the record.