Community Impact of Public Processions

2 Methodology

2.1 This report is based on data drawn from a multi-method approach which used both qualitative and quantitative data collection methods. The study consisted of three stages:

2.2 Stage One involved the collation of publicly available data provided by local authorities across Scotland on the number, range and characteristics of processions occurring routinely across Scotland (from 2010-2012). This statistical data was used to update the statistics originally provided in the Orr review (Orr, 2005) and to help inform the decision, taken in negotiation with the Scottish Government and project Research Advisory Group (RAG), as to which processions would form the focus for Stage Two. In most cases we were able to obtain the following information from local authorities:

• Procession dates
• Procession routes
• Any notable identified risks
• Principal organisations involved
• Event aims/purpose
• Notification decisions and reasons for any prohibitions or additional imposed conditions.

2.3 No information was returned from the following local authorities: Angus, East Dunbartonshire, Renfrewshire, Na h-Eileanan Siar, and Shetland Islands. In addition, some local authority areas were not able to supply a full three years' worth of data and/or were not able to supply data broken-down into relevant categories. For these local authority areas, the total figures were adjusted using the figures submitted for 2003 in the Orr Report (Orr, 2005)^[16].

2.4 Stage Two involved a retrospective study of community impact. Drawing on the information collated in Stage One, six processions that had taken place across Scotland were identified. Some of the factors that helped determine this selection included: attendance level (i.e. size); regularity (i.e. if annual event or not); level of policing required. From the outset, it was anticipated that selection would relate to the representative status of the events considered and proportionality.

Stage Two Site selection - Rationale and Process

2.5 Stage Two case-study sites were selected to provide a range of varied and prominent processions that could inform our collection of qualitative data, in terms of discussion with procession organisers and with statutory authorities. For instance we selected the main Boyne procession in Glasgow in July 2013 as this was a high profile event that had been subject to a great deal of scrutiny over the years by Glasgow City Council and the police. Conversely the Apprentice Boys of Derry procession in Perth was a rare example of a large Loyalist procession in a community that is does not traditionally host many Loyalist or Irish-Republican events. This allowed us to explore how the hosting of this event contrasted with the hosting of other community events in the area. Our discussions were by no means confined to a consideration of these processions, but they were intended to guide our initial approaches and conversations.

Table 2.1: Stage 2 processions

Processions selected for retrospective analysis	Date of Procession
Orange Order, Coatbridge	July 2012
Cairde na h'Eireann, Plains	October 2012
Royal Black Institution, Govan	August 2012
Scottish Defence League, Pollokshields	January 2013
Apprentice Boys of Derry, Perth	May 2013
Orange Order, Glasgow	July 2013

2.6 Semi-structured interviews were carried out with:

• procession/event organisers;
• local authority representatives/local councillors;
• senior police officers involved in strategic planning and the management of events.

2.7 Interviews sought to consider:

• the planning and preparation that preceded the event;
• the actual event itself - how it went, any disruption and if so, how this was addressed;
• consideration given before, during and after the event to its impact on the community/ies through which the event passed;
• the importance of such events in terms of cultural identity, traditional connection and community purpose which underpinned them.

2.8 Subject to receipt of the appropriate consent at the outset, interviews were digitally recorded and fully transcribed and coded for analysis. Research was conducted in accordance with the British Criminological Society Ethical Guidelines and subject to oversight by the University of Stirling, School of Applied Social Science Research Ethics Committee.

2.9 Stage Three involved a mixed methodological approach. Although initially intended to focus on four or five processions (see Group B in Table 2.2) which took place over summer 2013, we expanded this to include a wider range of processions where we undertook a more limited quantity of fieldwork (see Group A, Table 2.2)^[17]. Through meetings and discussions we generated ideas regarding a suitable sample of processions. It was apparent that six sites would not provide a systematically generalisable set of processions, but a series of case-studies that would have explanatory depth and from which generalisable lessons would need to be derived with caution.

Table 2.2: Stage 3 'live' events

Ethnographic data collection - GROUP A	Ethnographic plus residential, business, and on-street surveys - GROUP B
Scottish Defence League, (i)Pollokshields (ii)Central Glasgow, 27 July 2013	Orange Order Coatbridge, 6 July 2013 County Grand Lodge Procession
Pride Scotia, Glasgow Green 10 August 2013	Irish Republican Band Alliance, Airdrie 13 July 2013[18]
Royal Black Institution Renfrew, 10 August 2013	Cairde na h'Eireann Coatbridge, 13 July 2013
Scottish Defence League, Edinburgh, 17 August 2013	Royal Black Institution Glasgow, 10 August 2013 (i) Return parade to Bridgeton (ii)Return parade to Parkhead
Irish Republican 'Anti-Internment Procession with Prisoners' Families', Glasgow- 1 September, 2013	Pride of Govan Flute Band, Glasgow, 21 September 2013
Irish Republican Parade against Internment, Glasgow, 1 September, 2013

Stage Three Site Selection - Rationale and Process

2.10 Stage Three sites were selected to capture a range of public processions across several areas:

Coatbridge, North Lanarkshire: This is historically a community with a comparatively large Catholic population and where there have been tensions associated with Loyalist processions. The Grand Central Lodge procession was scheduled to take place in Coatbridge on the 6^th July 2013 and provided an opportunity to examine a very large procession (circa. 25,000 participants and spectators) through a relatively small town (circa. 40,000 residents)^[19]. In preparing for fieldwork in this area, we also became aware of an Irish Republican procession (involving a 'break away' Irish Republican band grouping) scheduled for the 13 July in the neighbouring town of Airdrie. As this was being policed by the same senior police officer, and Airdrie had recently hosted the Grand Central Lodge event, we observed this procession, and also undertook some street and business interviews. Also on the 13 July, Cairde na h'Eireann (Friends of Ireland) conducted their own annual procession in Coatbridge which we included in this study.

Grand Black Chapter (Royal Black Institution) procession, Glasgow: The Royal Black's 10^th August parade is the second largest Loyalist 'mass' procession after the Boyne parades, though it is to be noted that the Royal Black Institution does not attract as many spectators as Orange Order processions. The main Royal Black procession occurred in the small town of Renfrew, where we undertook some limited observations, though the focus of our fieldwork was on two smaller processions in the East End of Glasgow involving two Royal Black Preceptories (and associated bands) returning from the main Renfrew event. It was important to contrast the conduct and impact of larger processions with smaller chapter or lodge return processions. Earlier on the same day, Pride Scotia conducted their main gay pride procession in the nearby area of Glasgow Green which we also observed.

Scottish Defence League (SDL), Pollokshields/Glasgow City Centre/Edinburgh:

The increasing incidence of processions and static demonstrations organised by the Scottish Defence League (SDL) has been noted in recent years. These events have also been responsible for an increasing number of counter-demonstrations and have created a range of challenges for local authorities, Police Scotland and local communities. Although attendance at such events is relatively small, the presence of large counter-demonstrations and the potential for 'facilitated confrontation' raise a number of important issues. Following a controversial (and escorted^[20]) 'static' demonstration in the Pollokshields area of Glasgow in January 2013, a notification was made by the SDL for a procession in the same area on 27 July of that year. We chose this initially as a case-study site, though the subsequent refusal of the procession notification by Glasgow City Council led us to modify - but not abandon - our selection. Even though the procession did not go ahead, the notification itself was considered to have caused significant community anxiety, whilst anti-fascist organisers (Unite Against Fascism, UAF) determined to hold a static 'unity' event in the local area regardless. We therefore conducted interviews and focus groups with community and agency representatives, conducted observations of the static UAF event, and also observed the re-routed SDL event in Central Glasgow. Subsequent to this procession, the SDL were also given permission to conduct a procession in central Edinburgh, as was Unite Against Fascism. For comparative purposes, we conducted observations of these events.

Irish Republicans, Glasgow:

Two Irish Republican processions took place in Glasgow on 1 September. Both had a similar focus, to campaign against internment. The organisers of both processions (and the procession which took place in Airdrie) were viewed as 'dissident' Irish Republicans and their activities were considered controversial by the wider Irish Republican movement. The first procession, Free Ireland, was advertised as an Irish Republican Parade Against Internment and processed from Garnock St, Royston to the Gallowgate. The second procession, which took place later that day, was described as an 'Anti-Internment Procession with Prisoners Families' and processed from Jocelyn Square, Saltmarket to Janefield St via the Gallowgate. Structured and unstructured observations took place alongside a small number of business and on-street surveys.

2.11 Stage Three consisted of several methods of data collection:

• interviews with respondents (local authority representatives, police leads and procession organisers^[21]) who had a key role in the processions selected;
• ethnographic and structured observations of live processions;
• on-street mini-surveys of individuals proximate to live processions as well as short surveys of business premises near procession routes;
• face to face, postal and telephone interviews with local residents both prior to and after processions.

Residential surveys

2.12 Residential surveys contained detailed questions that aimed to assess the community impact of processions. On-street mini survey and business survey questions were also adapted from these question sets. The residential surveys had two phases. This allowed participants awareness and general attitudes to processions to be captured prior to the procession, and their experience of this specific procession to be captured after the event.

Pre-procession residential surveys

2.13 The first phase consisted of a short face to face survey conducted over the fortnight prior to the event. These interviews focused on:

• Awareness of the procession parade beforehand (how communicated);
• Any consultation on the procession or any feedback given;
• Any choices made as a result (e.g. avoidance);
• General attitude to processions and prior experience of processions;
• Views on how it feels to live in their local area including relevance/prevalence of discrimination in Scotland/their community and views on sectarianism.

2.14 The interviewers also captured basic community and demographic information. Respondents were asked to agree to participate in the second phase, and contact details were collected. Where respondents indicated they did not want to participate in a follow-up interview, one set of data was collected.

Post-procession residential surveys

2.15 The second phase involved a mixture of follow-up telephone interviews with each of the same households in the week following the procession, and separate post-procession surveys with a new sample of respondents. These interviews focused on:

• Whether respondents saw or heard the procession at any point;
• Their views on the conduct, meaning and purpose of the procession;
• If it caused them, or anyone they knew, any inconvenience, distress or harm immediately before, during or after the event;
• Whether they observed any stewarding or policing during the period of the procession, and whether they felt this was adequate and appropriate.

Achieved samples

2.16 Our final achieved survey numbers are show in Table 2.3. The original aspiration for the main pre and post residential surveys was to base this on a random, probability sampling approach, where address proximate to procession routes would be randomly selected from the Post Office Address File (PAF). The subsequent sample would then be visited in the two weeks before the procession, with respondents being interviewed on their doorstep. At this point respondents would also be recruited to take part in a post-procession telephone survey.

2.17 In the event this approach did not prove feasible. Our initial piloting of this approach in Coatbridge struggled with low response rates which were not primarily due to refusals but an inability to contact householders. Given that the survey was conducted in the first weeks of the summer holidays this may have been unavoidable. Our observational research of the 6^th July procession also led to the identification of a very distinct housing sub-area that had been missed when drawing a random sample of addresses. This sub-area, (a small network of streets around Gartsherrie road) was subsequently surveyed on a convenience basis after the procession.

Table 2.3 Survey sample sizes and response rates

Survey type	Areas	Target sample	Achieved sample	Response rate	Refusals
Pre-procession surveys
Pre-procession residents, door-to-door	Coatbridge	335	47	14%
Pre-procession, on-street surveys	Parkhead, Govan	n/a	131	n/a (convenience sample)	39%
Post-procession surveys
Post-procession telephone surveys	Coatbridge, Parkhead,Govan	108 pre-procession respondents agree to follow-up	62 are successfully re-contacted & interviewed	57% (35% of all pre-procession surveys)
Post-procession postal surveys	Coatbridge, Parkhead, Govan, Bridgeton	2000	192	9.6%
Post-procession, door-to-door	Coatbridge, Gartsherrie Road area.	n/a	39	n/a (convenience sample)	23% (where someone at home)

2.18 Subsequent attempts to roll-out our pre-post survey design in other case study areas proved even more impractical, as these other areas (Govan, Parkhead and Bridgeton) included a very high number of addresses contained within multi-address residential blocks which could only be accessed via controlled entry buzzer systems. These systems were frequently inoperable, and in other instances rarely generated response from householders. The areas were also characterised by high rates of social deprivation, with a large number of vacant (and indeed demolished) buildings. We therefore switched to convenience sampling for our pre-procession sampling, conducting on-street surveys in busy areas within two case-study sites (Parkhead and Govan), excluding stopped individuals who were not resident in the surrounding area.

2.19 Given the difficulties with the original pre-procession survey design, the post-procession element was adapted. Though the post-procession telephone follow-up was retained with reasonable success in terms of response rates, the value of this data proved questionable in practice. This element was predicated on the notion that as a result of the impact of one procession in their community, some detectable change in residents' attitudes might be observed. In the event, the fact that no changes were observable even at the level of the most basic descriptive statistics did not prove surprising. With experience in the field it become obvious to the research team that residents - who in most case had lived in areas for many years, and where those areas typically hosted a dozen or more processions every summer - were unlikely to have their views altered by yet one more procession. This does not preclude of course, the possibility that processions may not have a cumulative impact on residents (see our discussion in Chapter 7), merely that trying to detect impact on the basis of one procession proved naïve. Therefore, unless otherwise stated, all the analyses of pre and post survey samples in the findings chapters relate to separate, independent samples.

2.20 Residential survey responses were boosted by post-procession postal surveys. Here surveys were sent to addresses, again selected randomly from PAF. However, response rates for this element of the survey proved low, with the Royal Mail returning many surveys marked as 'address inaccessible' or 'addressee gone away'. Moreover, given the low response rate, and given that responses to the postal surveys proved markedly more negative than other survey responses, it seems highly likely that there was an element of self-selection bias. It may be that residents who were annoyed or upset by processions chose to return the survey.

2.21 The majority of survey responses collected were therefore, based on convenience sampling approaches, and this is a limitation that must be borne in mind when reviewing the findings. Statistical findings should be treated as indicative only, and not generalizable to a wider population. Conversely, it should also be noted that there was nevertheless a strong consistency in the pattern of findings across the different areas and across the different survey types and survey samples. Table AA1 in Annex A shows how representative the demographics of the achieved samples were relative to the population demographics of the areas in which the surveys were conducted. This shows that respondents were drawn from a range of age groups (although with a higher proportion drawn from the 45-64 age group than the 16-29 - a pattern particularly pronounced within the postal survey) and that the sample was roughly split according to gender^[22]. Respondents also demonstrated a generous and high level of engagement with the research process, providing valuable quantities of additional information. We therefore have confidence that in spite of the logistical difficulties encountered, high quality data capturing an indicatively useful range of views from a range of hard-to-reach populations was elicited.

On-street 'mini-surveys' and business surveys

2.22 On the day of processions mini-surveys were carried out on-street across four sites (Coatbridge, Airdrie, Govan and Gallowgate), with 138 conducted in total. Responses related to either a Loyalist or an Irish Republican procession (one exception being a respondent in the Gallowgate who had observed a Pride Scotia procession).

Respondents were surveyed in relation to the following processions:

Coatbridge (after the Orange Order procession)	6th July 2013	40
Coatbridge (after the Cairde na h'Eireann procession)	13th July 2013	27
Airdrie (after the Irish Republican Bands Scotland procession)	13th July 2013	28
Gallowgate (after the Royal Black return procession to Bridgeton)	10th August 2013	21
Govan (after the Pride of Govan Flute Band procession)	21st Sept 2013	16
Gallowgate (after the Anti-Internment procession)	1 Sept 2013	5

2.23 The sample cannot be characterised as a random sample of the local population. However, our approach to sampling can be viewed as reasonably systematic and justifiable in so far as we tried to sample all individuals passing nearby a procession route shortly after a procession. Our population of interest was people that may have encountered the procession, whether deliberately or by accident. One caveat to the systematic nature of our approach was that fieldworkers were steered away from approaching people that were laden with heavy shopping bags, people in procession or band uniforms, parents with small crying children or people who were obviously drunk. Moreover, whilst, 'vocal' refusals could be noted by fieldworkers, it is quite probable that many refusals were unobserved (e.g. people taking avoidance behaviour by crossing the road to avoid being stopped by the fieldworker).

2.24 On the day of processions we also convenience sampled businesses proximate to survey routes (there were few refusals and researchers attempted to survey all available businesses, but there was also no reliable sample frame and some businesses were shut).^[23] In total 104 businesses were surveyed.

Participant and non-participant observation

2.25 For each 'live' event, two researchers were embedded in the procession, one walking along the route on its periphery; and the other who undertook a visual and narrative tour of the procession, walking alongside a representative from the organising group where possible. This approach provided an opportunity to examine the procession from vantage points of 'outside looking in' and from 'inside looking out'. Both researchers were able to note the visible and audible events that characterised the procession, and importantly, to detect the emotional and experiential impact of the procession on participants and spectators. Each researcher observed and recorded the procession from their distinct viewpoint, noting interactions between the participants, general public, stewards and police. In addition to this 'unstructured' observation, a team of researchers conducted 'structured' observations at key points along the procession route, moving alongside the procession and pausing at key strategic locations to note qualitative impressions of the behaviour of participants and spectators, police deployment, interactions between participants, spectators and police officers; and to record songs and chants, as well as the general environment through which the procession moved.

Combining data sources

2.26 Utilising qualitative and quantitative methodologies and incorporating an ethnographic approach during the events provided an opportunity to synthesise and verify data collection and analysis through a process of triangulation. The combination of retrospective data collection and analysis and participation in 'live' events captured using ethnographic methodological approaches as processions took place, provided an important opportunity to capture the emotional context and atmosphere of the processions and experiential accounts of participation, observation and interactions. Full and further details of the research methods used and how residents and other participants were sampled can be found in Annex A.