Challenges around collecting equality evidence
Collecting regular and consistent information broken down by equality characteristics has been a challenge, even at a national level, in Scotland for a number of years. Reliable data on gender and age is available to local level for a wide range of variables. However, information on disability can be more challenging, and obtaining information on ethnicity, religion and sexual orientation is often not possible due to the small population numbers which are involved.
In the 2011 Census, minority ethnic groups made up 4% overall of the Scottish population. The Asian ethinc group made up more than a third of this proportion, indicating the diversity of the population in Scotland. In addition, a large proportion of these populations wre concentrated in a relatively small geographic area. Data on non-Christian religions revealed similar small numbers and geographic considerations. Further, the most recent data we have from the Integrated Household Survey estimates the total Lesbian, Gay and Bisexual population of Scotland to be 1.3%. Overall therefore, it is clear that 'small numbers' and 'small samples' characterise these populations in Scotland.
To achieve sample sizes which would be robust and large enough for thorough analysis from the household surveys would require the overall population sample to be very large indeed. This would incur significant practical and cost implications. More targeted sampling strategies and boosts can be effective but have disadvantages in terms of complexity of sample design, less precision about results, significant cost and not having an up-to-date list of names and addresses from which to draw a sample. Also, targeting specific groups might mean that the same people are surveyed time and again.
Often we need to know more than just the number of people who are in a particular group. For example, we might want to know what percentage of Pakistani people live in private rented accommodation compared to Indians, or how do views of local neighbourhoods differ between Muslims and Christians. To be able to answer this type of question we need a much larger sample of Pakistanis or Muslims than is necessary to estimate how many Pakistanis or Muslims there are in Scotland. As a result, analysis of ethnic and religious groups from survey based data is often limited to comparing “all ethnic minorities” or “all non-Christian groups” which treats all non–White ethnic groups and all non-Christian groups as the same, and hides important differences between different groups.
Up until recently, sexual orientation has not been asked on most surveys in Scotland. However new questions have been developed and are now being used in surveys, but once again, the small number of people responding will limit the utility of the data. Please see the guidance note on the recommended question to collect information on sexual orientation.