New measures of public sector output
Andrew MortimerShort-term Economic Indicators Branch
• Background– Why new measures are required– implementation
• Public administration and defence– Civil service and local authorities– Defence– Public order and safety
• Education– Public sector (to be covered later)– Further education– Higher education
• Health & Social work– Health– Adult social services– Children’s social services
Background
GDP – weights and seriesSector 2003 Weight
(percent)Number of Series
Agriculture etc 2% 7
Oil & Mining 1% 4
Manufacturing 15% 144
Energy & Water 3% 3
Construction 7% 11
Services 50% 78
Public Services 22% 19
Total 100% 266
Historical context• Previous assumption of non-market output = inputs• System of National Accounts 1993 / European
System of National Accounts 1995• Eurostat price and volume handbook (2001)
– ‘A’ methods: multiple direct measures of output, quality adjusted
– ‘B’ methods: Output indicator, not quality adjusted; or multiple quality adjusted input measures for collective services
– ‘C’ methods: Use of single input volume indicators (e.g. employment)
• ONS response to ESA95 – ahead of the game• Atkinson review
Actions• Supply data to ONS for UK National
Accounts.• Replicate productivity analysis for Scotland.• Ensure UK assumptions make sense for
Scotland.• Consider inclusion in Scottish GDP system.• Consider policy relevance of new measures
24th October• New economic target makes Scotland/UK
comparability ESSENTIAL!• Will introduce as many new measures as possible
into the next release of quarterly GDP– Ideally quality-adjusted output– Implement simple output measures if they are
similar to UK– May consider revised input-based methods if they
are an improvement over current estimates.• Improve scope for further incremental improvements.
Public Administration and Defence
Public Administration and Defence
7.2% of Scottish GVA Current measures Proposed measures
Civil Servants – numbers employed FTE employment in Civil Service by grade, weighted by salary.
---DWP also included---
Armed Forces personnel based in Scotland. No change
Number of administrative and clerical staff in NHS
No change
FTE employees in Local Government FTE employment by grade, weighted by salary.
Scottish Police and related services numbers No change
New series: Direct output measures for justice services
FTE employment on Fire Services Direct output measures – numbers of fires attended, weighted cost by type; hours spent
on fire prevention activities; RTAs.
Changes in grade structure of the Scottish Government 1999-2007
1000 800 600 400 200 0 200 400 600 800 1000
A1/A2
A3
A4
B1
B2
B3
BF
C1
C2
C3
Others
SCS
1999 2007
Justice services• Currently not measured explicitly• New cost weighted activity measures:
– Number of convictions:• High Court• Sheriff Solemn• Sheriff Summary• District
– Social work orders:• Probation Orders• Community Service Orders• Supervised Attendance Orders• Drug Treatment & Testing Orders• Restriction of Liberty Orders• Social Enquiry Reports
– Prisons:• Average daily prison populations
Fire services• Currently using staff numbers• New cost weighted activity measures:
– Fire response • 51 distinct measures• Type of residence / industry
– Special services• Road traffic accident• Other
– Fire prevention• No comparable Scottish measures available at this time (number of
hours engaged in fire prevention and community safety)• Low weight• Use UK data in the interim
Education
Education 6.5% of Scottish GVA
Current measures Proposed measures
Primary school teachers/
Secondary school teachers
Direct measure of output: number of pupils, adjusted by attendance and quality
adjustment for changes in attainment.
New measures: special education, teacher training and nursery education.
LA – other education employees Delete: included in output measures
Number of grant-aided and independent teachers
Number of students in independent schools.
Number of teachers in FE colleges Number of students (hours attended), cost weighted by type of course
Number of University lecturers. Number of students in higher education.
Further Education
• Currently Using Staff Numbers.• Further Education Funding Council PIs
– Student Units of Measurement (SUMs)• 40 Hours of Staff student contact.
– Cost Weightings (13 broad course types) are also available.
– Further PIs allow for future quality adjustment • Success and retention measures which depend on nature of
course. (e.g. assessed or non-assessed.)
• Comparable to current UK approach (but better).
Further Education
80
85
90
95
100
105
1999 2000 2001 2002 2003 2004 2005 2006
Current measure Proposed measure
Higher Education• Currently Using Staff Numbers• Higher Education Statistics Agency
– FTE Student Numbers are available and propose to use.– Complete alignment with existing UK measures
• Future refinements:– Quality Adjustments possible in principle but will require further
development– Candidate measures for research activity are less obvious.
Potentially:• numbers of PhD’s awarded; or• deflated research grants and contracts.
Health and Social Work
Health6.2% of Scottish GVA
Current measures Proposed measures
Health: Employees in Employment: NHS
To be determined. Possibly change to deflated expenditure whilst a
direct measure of quality adjusted output is finalised.
GPs Included in above
Dentists Included in above
Private and other health No change.
Veterinary practices No change.
Health services
• Currently use 3 employment series– NHS staff; GP’s and Dentists
• Direct measure of output still in development
• Alternative approach: deflated inputs– 74 distinct measures (staff and supplies)– Deflated at source using unit costs– If used, will lead to minimum change to existing estimates
• Final decision on approach to be confirmed
Social services6.2% of Scottish GVA
Current measures Proposed measures
FTE Employees in LA Social Work Activities
New direct measures of output for adult and children’s services; number
of bed-nights in residential care, number of hours in day care and
domiciliary care, number of referrals etc.
Non-LA employees in Social Work activities
Ditto.
Social services• 10 new direct output measures
– Adults• Care Homes (resident weeks per year)• Home Care (hours per year)• Day Care (sessions per year)
– Children• Secure Accommodation (child-days)• Residential Schools (number of pupils)• Care Homes (number of children)• Adoption Services (number of children)• Fostering/Family Placements (number of children)• Assessment (deflated inputs)• Other Services (deflated inputs)
• Weighted by relative cost of delivering each service
Conclusions
Conclusions
• Significant methodological improvements to be introduced in next GDP release
• Improved comparability with UK methods• Greater scope for future incremental improvements• Too early to gauge full impact
• Further work over next 12 months on:– Direct output measures for Health services– Replication of UK measures of social security output– Quality adjustment of Higher & further education– Replication of future UK developments for Scotland
Discussion
New measures of public sector output
Reserve slides
Atkinson Review - History
• 2003 - No 10 ! • Tony Atkinson – 1 Year’s commission• Interim Report – July 2004• Final Report – January 2005• 54 recommendations to improve the
measurement of public services output and productivity
• Implementation Strategy for Scotland – April 05• Action Plans – updated quarterly• Working teams set up in Scottish Executive.• Established links with ONS UKCeMGA.
Atkinson: summary of key principles
• Principle A: the measurement of government non-market output should aim to follow a procedure parallel to that adopted for market output.
• Principle B: the output of the government sector should in principle be measured in a way that is adjusted for quality of the service delivered.
• Principle C: account should be taken of the complementarity between public and private output – e.g. NHS output increasing private sector productivity.
• Principle E: measures should cover the whole of the United Kingdom; and reflect regional variation in the delivery of public services and/or data collection.
• Principle F: the measurement of inputs should be as comprehensive as possible, and in particular should include capital services – e.g. imputed rent for buildings
• Principle G: Pay and price deflators should be sufficiently disaggregated to take account of changes in the mix of inputs; and should reflect full and actual costs.
Atkinson RecommendationsInput (proxy) Poor
Total Inputs - deflated Better
Output (proxy) Even Better
Total quality-adjusted Output - deflated
Good!