Scotland's New Ethnicity Classification for Scottish Official Statistics and Recommended for Scotland's 2011 Census

Annex E Proxy Comparisons Between the New Classification and 2001 Census Classification

How can I compare data from the old and new classifications?

1. Many data users will want to compare ethnicity data derived from the old classification (2001 Census) and the new classification, to produce trends over time. Because it was necessary to modernise and change the new classification this means that direct comparability is not possible. However, many aspects of the classification are similar and this means that classifications are broadly or indirectly comparable i.e. allow proxy comparisons.

2. The table above shows the old and new categories and tick boxes side by side. The comparisons that will and will not be possible (in some cases) are explained below, together with any necessary caveats (or exceptions). It is not possible to predict exactly how different people will answer the new classification, however it is possible to give an indication of what is likely to happen based on knowledge gathered during the course of this review. It is strongly advised that data users who produce comparisons over time consider these issues in any analysis they produce.

3. ' White' Category. At category level, total counts of 'White' should be broadly comparable. However, some people identifying as 'Gypsy/Travellers', 'Polish' and to some extent other people from Central and Eastern European countries may have responded in the previous classification using the 'Other ethnic background' category. At tick box level only the 'Scottish' and 'Irish' tick boxes are broadly comparable, although some people identifying as 'Scottish' under the previous classification may identify as 'British' under the new classification. The 'Other British' (2001) tick box is roughly comparable with the sum of the new 'English', 'Welsh', 'Northern', 'British' and 'Gypsy/Traveller' tick boxes. The 'Any other White Background' (2001) is broadly comparable with the sum of the new 'Any other white ethnic group' and 'Polish' tick box.

4. 'Mixed or multiple Ethnic Groups' category. Total counts of this category are broadly comparable. However, some people identifying as, say, 'Indian Scottish' or 'African British' (who answered using the 'Mixed' category in the previous classification) may now answer using the new 'Asian, Asian Scottish or Asian British' tick boxes or the 'African, Caribbean or British' tick boxes.

5. 'Asian, Asian Scottish or Asian British'. At category level, total counts should be broadly comparable. However, as detailed above, we may expect some respondents who responded using the 'Mixed' category in the previous classification to identity as, say, 'Indian Scottish' in this category using the new 'Indian, Indian Scottish or Indian British' tick box' for example. The same applies to all the 'Asian' groups listed in the new classification.

6. 'African, Caribbean or Black'. At category level, total counts should be broadly comparable. However as detailed above, some respondents who previously responded using the 'Mixed' category may now identify as, say, 'African Scottish' in this category using the new 'African, African Scottish or African British' tick box. At tick box level, comparisons are not advisable, since testing has indicated that some people will now opt to identify as 'Black, Black Scottish or Black British' rather than 'African' or 'Caribbean' as in the previous classification. It is possible that some people identify as 'White African' or 'White Caribbean' and may choose to do so under this category (although small scale testing indicated that the majority of those identify as 'White African' did so using the 'White' category on the new classification).

7. Changes affecting all the above categories. An analysis of responses from people identifying as 'Arab' in the 2001 Census, shows that 47 per cent did so under 'Any other ethnic background', 22 per cent under the 'Asian' category, 13 per cent under the 'Mixed' category, 7 per cent under the 'White' category, 1 per cent under the 'Black' category and 9 per cent gave multiple responses across several categories. The inclusion of the new 'Arab' tick box in the 'Other Ethnic Group' category means that, now, most of these people are likely to respond here instead. This will have reduce somewhat the total counts of each of the above categories.

8. 'Other Ethnic Group'. At category level, total counts should be broadly comparable. However, as detailed above, the inclusion of the new 'Arab' tick box should mean that more people identifying as 'Arab' use this category (with the possible exception of the people identifying as say 'Arab Scottish' or 'Arab British' who may continue to identify using the 'Mixed or Multiple Ethnic Group' category). As detailed above, we may expect fewer people from Central and Eastern Europe to identify using this category than did so under the previous classification, because the inclusion of the new 'Polish' tick box under the 'White' category is likely to encourage such respondents to give their answers here instead. The same is true of people identifying as 'Gypsy/Traveller', to some extent.