Hosted by: Funded by: Researching the Third Sector.

Ho

sted

by:

Fu

nd

ed b

y:

Researching the Third Sector

Guiding questions for session1. Representativeness: what are we not seeing?2. What are the challenges and opportunities provided by longitudinal perspectives?3. How can we identify and think about different dimensions of organisational change?4. Are specific methods needed for researching the third sector?5. The future: what are the next steps relating to data and method?

Ho

sted

by:

Fu

nd

ed b

y:

Reflecting on the use of administrative data for research: challenges, opportunities and future directions

John Mohan and David CliffordThird Sector Research Centre,

University of Southampton

Context: administrative data

• Exciting times for social science!• Flagship social surveys in UK, but increasingly

also access to administrative data• Here: reflecting on TSRC’s experience of use of

administrative data collected by Charity Commission

Context: Charity Commission data

• Charities – charitable purpose and public benefit

• Data collected as part of Charity Commission’s regulatory role

• Register computerised since early 1990s• Financial information (detail depending on

size), area of operation, what they do/ who they work with/ how they do it, etc.

Challenges

• What are we not seeing?• Differences in administrative practice

– Over time: changes in composition of charities on Register

– Over space: Charity Commission vs OSCR• Lack of metadata (but P-ADLS..)

Opportunities

1. Individual organisations’ income and expenditure over time

– Cross-sectional (snapshot) and longitudinal (following organisations through time) perspectives

2. Population of organisations (large numbers)– Detailed disaggregation of patterns possible: e.g. by organisation and area characteristics

Example of opportunities

• Example of work on survival of charities• Which kinds of charities, in which kinds of

areas, have a higher probability of dissolution?– Inactive organisations included in our data– Focus on those charities to have had at least

£5,000 in income at some point since 1995 (around 140,000 charities)

– Disaggregation by: initial income; type of charity; area

Probability of charity dissolution in a given year:charity and area characteristics

1

1.5

2

2.5

3

3.5

Dis

solu

tions

pe

r 10

0 ch

arit

y ye

ars

0102030405060708090100Least deprived Deprivation(percentiles) Most deprived

5k-10k 10k-100k 100k+

Probability of charity dissolution in a given year:by local authority

0

1

2

3

4

Dis

solu

tions

pe

r 10

0 ch

arit

y ye

ars

050100150200250300Least deprived Deprivation(rank) Most deprived

Future directions..

1. Detailed financial information for a random sample of 10,000 charities (in collaboration with NCVO), e.g. income streams, income sources

2. Linking organisations (charities) and individuals (trustees)

Future directions..

3. Other sources of administrative data: move towards data sharing (‘open data’)-collected by voluntary organisations: finance and funding records; information about where services are delivered; feedback and performance data-collected by government and Big Lottery Fund: organisations and projects funded

Future directions..

Challenges of using these datasets-consistency-representativeness-metadata (P-ADLS: collating this information)-recognising limits as well as opportunities

Acknowledgements

• Steve Barnard, Peter Backus, Frida Geyne• Charity Commission• ADLS• National Council of Voluntary Organisations

For more information..

• www.tsrc.ac.uk– >Research>Quantitative Analysis