300 likes | 325 Views
This research utilizes administrative data from the Office of the Scottish Charity Regulator (OSCR) and the Milo database to analyze Scottish civil society organizations (CSOs). By comparing and cleaning data from both sources, this study aims to provide insights into the operations, categorization, and distribution of CSOs across different sectors and geographical areas in Scotland.
E N D
Using Administrative Data to Understand Civil Society Organisations in Scotland Dr Orian Brook & Dr Alasdair Rutherford University of Stirling
Measuring Civil Society Charity registers commonly used • Office of the Scottish Charity Regulator (OSCR) hold records on all registered Scottish charities (no income threshold) • c24,000 records, high quality and complete data (relatively) • But suffers Head Office bias: • legal address not operational address • Only one location per charity, no matter how many they have
Milo database Managed by Scottish Council for Voluntary Organisations, used by Third Sector Interfaces/Volunteer Centres across Scotland • Supporting CSOs in legal formation, training, advising trustees etc • Key activity: matching volunteers and volunteering opportunities Feeds various national online resources: • getinvolved.co.uk – search for third sector organisations • Volunteerscotland.net – search for active volunteering opps. • ALISS (A Local Information System for Scotland) – signposting to community support/resources • Goodhq.org (to be fully launched): share experiences of charities online (“tripadvisor for charities”) Therefore quality of data is increasingly important
Milo Data 35,000 voluntary orgs/38,000 orgs (plus volunteers, opportunities but not used for this project) Pros (compared to OSCR) Records working sites not legally registered addresses Records multiple sites (eg charity shops) Records vol orgs which aren’t registered charities Cons (compared to OSCR) Recorded for working not research – messy, inconsistent Incomplete – orgs not recorded if not interacting with TSIs/VCs
Cleaning Milo Missing data • c4,000 with no addresses • approx half of records have no “main activities” specified Duplicates – few identified, but more to discover? Out of date addresses – external cleaning (few) Charity Reg No. • Approx 10% of charity numbers didn’t match OSCR • By checking correct formatting, able to match all but 1%
Linking Milo to OSCR Charity Reg (cleaned) Deterministic – cleaned org name, postcode etc Probabilistic - % likelihood • Difficult to automate eg 103rd A, B, C and D City of Edinburgh Brownie Units all separate registered charities Still only matched 38% of Milo records, matching to 48% of OSCR Hoping further work might increase this but still likely to be about half of each database not overlapping with the other
Coding Milo & OSCR Using ICNPO (International Classification of Non-Profit Organisations) as widely-accepted for comparable studies Level 1 1 Culture and Recreation 2 Education and Research 3 Health 4 Social Services 5 Environment 6 Development and Housing 7 Law, Advocacy and Politics 8 Philanthropic Intermediaries, Vol. Promotion 9 International 10 Religion 11 Business/Professional Associations, Unions 12 Not Elsewhere Classified
ICNPO detail Level 2 codes also officially developed. Eg under social services: 4100. Social services (general) 4200. Emergency and Relief 4300. Income Support and Maintenance Level 3 codes more bespoke to SCVO To provide more discrimination, eg under 4100 4110. Services for Children and families 4120. Pre-school daycare 4130. Services for young people 4131. Scouts, Guides etc 4140. Services for people with disabilities 4150. Services for elderly people Or to specify CSOs which could belong to other level 1 codes, eg 2410 Medical Research Funds others,conducts research, provides health care & social support
Process On OSCR & Milo multi-select categories are inconsistently selected, v hard to analyse, often missing in Milo Inspired by Mohan/Barnard/NCVO work Coded charity registers (English and Scottish) Searched for keywords in charity name and/or aims Now using Milo, coding primarily by searching for specific keywords which accurately categorise CSOs Using code not manually so can be rerun for refreshed or different data Also referring to previous coding, to org aims, to other SCVO data and using generic keywords Coded all but 7% of records
Urban-Rural Classification Milo has higher proportion of CSOs outside Large Urban Areas – removal of head office bias? Milo records with address missing: as often recorded by TSIs/VCs within central belt as other records, so probably no major bias
Urban-Rural Classification Standardised by population shows greater impact of how Milo represents CSOs outside large cities
Indices of Multiple Deprivation Milo shows much higher proportion of CSOs based in more deprived areas, and fewer in least deprived areas
Indices of Multiple Deprivation • Analysis by population shows much more important presence of CSOs in deprived areas. • Address of CSO not = area of operation, but more likely on Milo, and unlikely to be based in a deprived area but not working in one
ICNPO Top Level Codes Relative distribution between lists mostly similar, differences not restricted to particular sectors
ICNPO Top Level Codes Milo has more CSOs in Culture & Recreation – consonant with intermediaries membership – good integration with TSI/VC orgs
ICNPO Top Level Codes Development & Housing closely aligned to civil society contribution to social policy – 44% of Milo records additional to OSCR
ICNPO Level 2 codes Social Services CSOs also closely aligned to social policy • Milo also holds 40% more records, and most subcodes have even higher proportion, with exception of Scout troops
ICNPO Level 2 Codes Milo holds more CSOs tackling mental health and addictions support, important for Scotland’s pressing public health issues
ICNPO Top Level Codes OSCR better represents Scouts, Religious orgs & grantmaking trusts – less integrated to TSI/VCs, no help needed for volunteers etc
Urban-Rural Classification Milo has higher proportion of CSOs outside Large Urban Areas – removal of head office bias?
Indices of Multiple Deprivation Milo shows much higher proportion of CSOs based in more deprived areas, and fewer in least deprived areas
Thank you! Orian.brook@stir.ac.uk