Scottish Prisoner Voting Survey 2022

3. Methodology

3.1. Questionnaire Design, Dissemination & Returns

The survey questionnaire was developed in collaboration with colleagues from the Scottish Prison Service and the Electoral Commission. The survey was designed to gather views of prisoners serving sentences of twelve months or less on a number of topics including, length of sentence (to determine eligibility), previous voting history, experience of registering to vote in prison and participation in the 2022 Local Government election. A free text response was also provided to allow prisoners to provide additional information.

Consideration was given as to whether the survey could also include prisoners on remand however a decision was taken against this as it was felt that questions targeted at prisoners would not be relevant or appropriate for those on remand.

As noted above the majority of the questionnaire was designed with closed questions to ease completion as well as analysis, with one voluntary free text response provided for prisoners to provide additional detail if they felt so inclined. The full list of questions can be found in the annex to this document.

With regards to distribution of the survey, the SPS research team distributed the survey to local prisons^[9] where prison officers were responsible for circulation within the prison itself in paper format to eligible prisoners. Prisoners were provided with an envelope within which they could place their completed survey, and which could be sealed before being handed back to the prison officer. Responses were then collected and sent back to the SPS research team who scanned the results and forwarded these on in batches to the SG research team for analysis. The responses were then input and coded within Microsoft Excel for analysis by the SG research team.

It was made clear to respondents that participation in the survey was both voluntary and anonymous and prisoners were provided with a privacy notice^[10] setting out how the data would be used as well as their data protection rights.

3.2. Response Rate & Approach to Analysis

The survey was launched in May 2022, shortly after the 2022 Local Government election had concluded and was open for responses until mid-June 2022. At the time the survey was launched, unpublished data received from the SPS demonstrated that there were 526 people in custody, who were over the age of 16 and serving sentences of twelve months or less who would be eligible to vote and therefore participate in the survey.

66 responses to the survey were received. 4 responses were removed as out-of-scope for the survey as either prisoners were serving sentences of longer than twelve months, or were awaiting sentence and therefore would not be eligible to vote or it would not be possible to determine whether they were eligible at the time of the survey. One duplicate response was also removed.

After these ineligible responses were removed, there were 61 responses to the survey, giving a response rate of 11.6%. Responses were received from 10 establishments. Just under half of all responses came from prisoners serving sentences within HMP Barlinnie. The latest Official Statistics on Scotland's prison population^[11] (for the financial year 2021-22) suggest that of the eligible prisoner population, 20% were serving sentences at HMP Barlinnie. This would suggest that prisoners from HMP Barlinnie were overrepresented. Conversely, HMPs Edinburgh and Glenochil were underrepresented, with no responses received.

It should be noted that SPS counts of prisoners are not directly comparable to the Official Statistics average prisoner numbers (more information is available in the associated technical manual^[12]); the above return and response rate are calculated using the SPS prisoner count.

Unique identifiers were applied to each prisoner to aid analysis. Some prisoners chose not to answer all questions, the analysis therefore only includes valid responses for each question and therefore a different base number may exist per variable.

With regards to how the data has been analysed and presented within this report it should be noted that the findings have not been weighted or had confidence intervals applied, nor has any statistical significance testing been conducted. There is therefore an unknown margin of error associated with each result.