We have published a full report called Sexual Orientation in the Census that evaluates the results of a small-scale postal survey including a question on sexual orientation.
A case was put forward for a question on sexual orientation to be included in the 2011 Census. A primary driver for this was the need for data to baseline the monitoring of equality legislation which was in effect by then.
We conducted a small-scale postal survey as a first step to understand public attitudes to a sexual orientation census question and the feasibility of including it.
4,400 individual census style forms were sent to a geographically representative random sample of Scottish households. There were four variants of the form which are explained in paragraph 3.1.3 of the report. Half of the forms included a sexual orientation question, half did not.
There was no difference in the response rate between form variants. However the overall response rate for the survey was only 31 per cent and certain groups were more likely to respond than others. This result cannot be generalised to draw conclusions about the response rate of the public as a whole.
The results from the sexual orientation question are summarised in table 4 of the report. Overall only 2.2 per cent of respondents declared non-heterosexual sexual orientation. The non-response to the question was around 6 per cent and a further 8.5 per cent of respondents selected “Prefer not to answer”. Therefore in this test the percentage of respondents who did not provide useful data is around 14 per cent. This far outweighs the percentage of respondents who declared a non-heterosexual sexual orientation. The survey results call into question the accuracy of data gathered by such a question and hence the utility of any such data.
Download Sexual Orientation in the Census (342Kb pdf file) for a full report.