UK Centre for Measurement of Government Activity Public Service Output, Input and Productivity: Education

Transcription

1 UK Centre for Measurement of Government Activity Public Service Output, Input and Productivity: Education Richard Wild Fraser Munro Daniel Ayoubkhani

2 UKCeMGA: Public Service Output, Input and Productivity: Education 1. Productivity of Education in the UK This is the third ONS education productivity article, which updates the estimates of change in productivity of publicly funded education services published in Total Public Service Output and Productivity in June 2009 (ONS 2009a) and gives the first estimate for Latest estimates of productivity show that: productivity was the same in 2008 as in 1996; this is because over the whole period: output grew by 33.4 per cent, with an annual average increase of 2.4 per cent inputs grew by 33.3 per cent, with an annual average increase of 2.4 per cent However, within the whole period: from 1996 to 2000, productivity grew by 8.1 per cent, with an annual average increase of 2.0 per cent, driven by relatively strong growth in the school age population coupled to relatively weak growth in inputs from 2000 to 2005 productivity fell by 6.8 per cent, an annual average fall of 1.4 per cent. This resulted from a steady rise in the number of pupils attending school, once adjusted for quality, being outstripped by a sharp rise in inputs, mainly through the employment of more school support staff from 2005 to 2008, productivity was fairly flat and fell on average by 0.2 per cent, as output and inputs grew at a similar pace. Between 2007 and 2008, output and inputs grew by 2.0 and 2.7 per cent respectively, causing productivity to fall by 0.7 per cent Figure 1.1 Growth in education output, inputs and productivity estimates, Percentage change from Output Inputs Productivity Per cent Annual average percentage change Volume of output Volume of inputs Productivity

3 1.1 The pattern of productivity change in the UK is dominated by changes in England. The pattern reflects: Background changes in the secondary-school-age population in the UK, which increased from 1996 to 2004 then declined from 2005 to 2008, and changes in the UK primary school age population, which fell sharply from 1998 to These changes reduced education quantity. Where it was not locally feasible to close schools or reduce the number of classes, measured productivity will have fallen if fixed and semi-fixed costs were spread over fewer pupils improved attendance rates in the UK from 1996 to 2008 and improvements in GCSE and equivalent and Standard Grades. These changes increased education output as they more than offset the decline in the school age population. However, the improvements have required some targeted resources a large increase in the number of support staff between 1996 and 2008, which in recent years was likely influenced by the agreement in 2003 of a work reform package with the UK and Welsh governments, employers organisations and unions, intended to reduce teacher workload. This change has caused inputs to rise Rapid expansion in both publicly funded nursery provision and health professional students. Publicly funded nursery provision increased following the introduction of policy in England providing a number of hours of free childcare per week for three- and four-year-olds, while the rise in health professional students has been driven mainly by trainee nurses, particularly at degree level It may also reflect: improvements in the quality of education, through increases in support staff numbers, in ways that are not currently measured by the quality adjustment. For example, through support helping the integration of pupils with special needs expenditure on initiatives such as Every Child Matters 1 and narrowing the gap 2 that are currently not reflected in the output measure but increase the volume of inputs time lags, which mean that some changes in resources, including expansion in pre-school education, may not yet have had any impact on the current quality measure 1.2 This article continues work to incorporate recommendations of the Atkinson Review: Final Report Measurement of Government Output and Productivity for the National Accounts (Atkinson 2005) and subsequent methodological developments endorsed in Measuring Quality as Part of Public Service Output (ONS 2007a). 1.3 This article presents estimates of multi-factor productivity growth, equal to the growth in the ratio of the volume of output to the volume of inputs. The estimates presented in this article reflect those presented in Total Public Service Output and Productivity (ONS 2009a) from 1997 to 2007 in terms of their construction. However, this article presents more detailed analysis and provides estimates from 1996 to 2008 using the latest data. 1.4 It is unlikely that a single measure of productivity change will ever capture all the costs and benefits of education. The methods give approximations to a complex reality and estimates therefore need to be interpreted carefully (see section 4). 1.5 While the methods and data for this article are based on the National Accounts, there are some adjustments and developments unique to this article, which are explained. 1.6 In order to provide an indication of reliability, revisions since the last published ONS estimates of education output, inputs and productivity figures (ONS 2009a) are presented in Annex A. 1 Details available at: 2 Details available at: 3

4 1.7 In the rest of this article: section 2 reports the latest estimates of change in education output section 3 reports the latest estimates of change in the volume of education inputs section 4 provides further contextual information section 5 summarises future development plans 2. Output of Education in the UK Latest estimates of output show that between 1996 and 2008: output grew by 33.4 per cent, with an average annual increase of 2.4 per cent, and output grew in every year growth was strongest from 1996 to 1998, averaging 3.3 per cent annually the rise in output is driven by changes in pupil numbers and attendances, which are largely driven by demographic patterns, and increases in GCSE attainment Figure 2.1 Growth in education output, Percentage change from Per cent Annual average percentage change Volume of output

5 2.1 Education quantity Education output has two components: Quantity (which is adjusted by) Quality This section considers the quantity component In the context of the National Accounts and this article, education quantity 3 includes: attendance-adjusted pupil numbers at government-maintained schools full-time equivalent places in pre-school education (including publicly funded places provided by the private, voluntary and independent sector) the number of students on initial teacher training and health professional courses funded directly by government The National Accounts measure of education quantity, in accordance with international guidelines, does not cover further education (which includes sixth form colleges) or higher education, since these are funded through grants and transfers and are therefore classified outside the public sector. However, a large proportion of further education funding comes indirectly from the public sector and so there is a logical case for including any such expenditure alongside directly funded education activities. This article includes figures for those aged under 19 on publicly funded further education courses, as these students complement those already measured in school sixth forms. Future work is planned to include the remainder of publicly funded further education and higher education in the productivity measure (see section 5) The quantity measure for each of the components of education output is based on pupil or student numbers in each academic year. Pupil numbers in primary, secondary and special schools throughout the UK, and city technology colleges (CTCs) and academies in England are adjusted for attendance in order to reflect more accurately the quantity of education services delivered. As attendance at these institutions is compulsory, changes over time are largely driven by demographic patterns, with the exception of the relatively recently established CTCs and academies No attendance adjustments are currently made to the quantity of UK pre-school education, initial teacher training (ITT), health professional training and further education (FE) as these are not part of compulsory education Figure 2.2 summarises the quantity indices for each education component over the period 1996 to Figure 2.2 Quantity measure of education, =100 Average annual percentage change Pre-schools Primary Schools Secondary Schools Special Schools CTC / Academies Initial Teacher Training Health Professional Training Further Education for U19s Total Education quantity, defined here as the sum of publicly funded educational services delivered prior to the application of any quality adjustments, can be measured in terms of pupil and student numbers within a specified time period. The Eurostat Handbook on Price and Volume Measures in National Accounts (Eurostat 2001) recommends that education output, except in further and higher education, should be measured by the number of pupil hours. The number of students is considered a better reflection of output in further and higher education. 5

6 2.1.7 Education quantity grew by 4.6 per cent between 1996 and 2008, with an average annual increase of 0.4 per cent. The largest observed increase was in England-only CTCs & academies which grew by per cent. This was due to a rapid policy-driven expansion in academies (some of which were previously CTCs), although most of this increase came from the conversion of existing secondary schools. Health professional training grew rapidly, by per cent, mainly through an increase in the number of trainee nurses, particularly at degree level. Pre-school education also grew rapidly, by 70.8 per cent, due to policy introduced in England providing free childcare for all three- and four-yearolds for a set number of hours per week. Falls in quantity were observed in primary schools, due to falls in the five- to eleven-year-old population, and special schools, due to policy that integrated many children with special needs into mainstream schools. Further education for under 19s grew by 24.4 per cent over the period, with much of the rise coming from students taking qualifications suitable for entry into higher education. The quantity of ITT also fell, primarily due to a shift in demand from threeto one-year courses which reduced the number of students Figure 2.3 summarises the relative shares of expenditure on each of the education components over the period 1996 to Figure 2.3 Expenditure-based education weights, Per cent Pre-schools Primary Schools Secondary Schools Special Schools CTC / Academies Initial Teacher Training Health Professional Training Further Education for U19s Total Primary and secondary schools constitute the largest proportion of expenditure, accounting for around 77 per cent of the total in Between 1996 and 2008, the proportion of expenditure on primary and special schools fell as the school-age population and the number of pupils fell. Conversely, the proportion of expenditure on preschools and secondary schools rose, as the number of children attending pre-schools and secondary schools (including CTCs and academies) increased. The relative proportions of expenditure on other components remained largely unchanged Figure 2.4 illustrates contributions to total annual quantity growth by component of education in terms of percentage points over the period 1997 to Between 1997 and 2008 education quantity grew by 4.6 per cent. This growth was driven by positive contributions of 2.3, 1.7, 1.6, 1.6 and 1.2 percentage points from secondary schools, further education for under-19s, health professional training, pre-schools and CTCs and academies respectively. These were partially offset by negative contributions of 3.1, 0.5 and 0.2 percentage points from primary schools, special schools and ITT respectively. CTCs and academies have made an increasing positive contribution to education quantity growth since academies were established in 2000, although this is largely a substitution effect in terms of the quantity of the secondary education they replace. 6

7 Figure 2.4 Contributions to education quantity growth, Percentage points Secondary Schools Further Education Health Professional Training Pre-school CTC / Academies Initial Teacher Training Special Schools Primary School Secondary Schools Further Education Health Professional Training Pre-school CTC / Academies Initial Teacher Training Special Schools Primary School Education quality Percentage points The Atkinson Review and the Eurostat Handbook on Price and Volume Measures in National Accounts (Eurostat 2001) recommend that public service output should be measured in a way that adjusts for quality change. Quality adjustments are currently applied to elements of compulsory schooling and ITT, while further education, health professional training, special schools and preschools are not currently adjusted In 2009, ONS conducted research into the use of Key Stage 4 test results as a way of representing quality change in compulsory schooling by school level (ONS 2009b), in line with Recommendation 9.3 of the Atkinson Review (Atkinson 2005). Due to the practical limitations of the proposed methods, the ultimate conclusion of the development work was that ONS should continue to quality-adjust the output of school education using the existing methodology based on the annual change in GCSE and equivalent average point scores (APS). Further details can be found in Methods Development: Decomposing Quality Change in the Output of Publicly Funded School Education (ONS 2009b) The current schools adjustment uses the change in the uncapped APS of GCSE and equivalent qualifications in England and Wales and Standard Grades and equivalent qualifications in Scotland to estimate quality change for government-maintained primary and secondary schools and CTCs and academies The APS for GCSE and equivalent qualifications relates to the attainment of pupils at the end of Year 11 (pupils aged 15 16). It is assumed that the change in the APS, used to approximate quality: Total 4 Key Stages 1 to 4 relate to primary and secondary pupil assessments in England, Wales and Northern Ireland, although they are applied differently in the national curricula. The research used England data only, for Key Stage 2 tests taken at age in Year 6 (end-primary) and Key Stage 4 tests (GCSEs and equivalents), taken at age in Year 11 (end-secondary) 7

8 should be applied to all pupils in primary and secondary schools (from reception class to the end of the sixth form) in the UK and CTCs and academies in England is an adequate approximation for all education outcomes, for example attainment after 16 and development of wider outcomes such as citizenship is the best measure for the annual change in the quality of output, based on current research This quality method is not the same as the one used for schools in the National Accounts. An outline of the National Accounts method can be found in Methods for public service productivity: quality adjusting school education output, September 2007 (ONS 2007b) A time series of GCSE scores for Northern Ireland is not readily available so the change in APS in Northern Ireland has been approximated by the change in APS in England. Further information on the quality adjustment for schools output can be found in Public Service Productivity: Education (ONS 2007c) In addition to compulsory schooling, ITT is also adjusted for quality using the change in the proportion of final-year students who attain Qualified Teacher Status (QTS) in England. Data are available from 2001/02 and these have been used to construct a quality adjustment that is applied to the whole of the UK. 2.3 Impact of quality-adjustments on education quantity Figure 2.5 illustrates the impact of quality-adjustments on education quantity over the period 1996 to Figure 2.5 Impact of quality-adjustments on education quantity, Percentage change from Unadjusted output (quantity) Quality adjusted output Per cent Average annual percentage change Quality adjusted output Unadjusted output (quantity) When quality adjustments are applied to education quantity, the overall growth over the period 1996 to 2008 increases from 4.6 to 33.4 per cent, with the average annual percentage change increasing 8

9 from 0.4 to 2.4 per cent. This difference illustrates the importance of applying measures that capture quality change fully An extended analysis of education output, including demographic information and country-level data, is presented in the complementary article Public Service Output, Input and Productivity: Extended Analysis of Output (ONS 2009c) Work is ongoing to develop a quality measure for pre-school places. Future work will look into the development of measures for further education and health professional courses, in line with the recommendations of the Atkinson Review (Atkinson 2005). The use of annual changes in GCSE/Standard Grade APS to adjust school quantity will also continue to be reviewed to take account of emerging research on the contribution of education to pupil and student outcomes. Potential improvements to the schools quality adjustment methodology are presented in Annex B. 3. Inputs to Education in the UK Latest estimates of inputs show that: between 1996 and 2008, inputs grew by 33.3 per cent, with an annual average increase of 2.4 per cent between 1996 and 1997, inputs fell by 0.3 per cent, but grew in all other years between 2000 and 2002, inputs grew strongly, by 9.7 per cent between 2007 and 2008, inputs grew by 2.7 per cent A key driver of the rise in inputs has been increases in support staff numbers. This rise may lead to changes in the quality of education that are not currently captured by the existing quality adjustment, which is based on GCSE and equivalent results Figure 3.1 Growth in the volume of education inputs, Percentage change from Per cent Annual average percentage change Inputs

10 3.1 Components of education inputs at current prices Education inputs have three components: labour (for example teaching staff), goods and services (such as learning materials and electricity), and capital services (for example the flow of services provided by a vehicle or building in a given period) Expenditure on labour and goods and services is measured in current prices (what was actually paid). Figures for capital services are estimates of the value of the flow of services from education capital; while they do not form an explicit part of publicly funded education expenditure, they represent the annual input provided by capital assets owned and are therefore included alongside actual current expenditure Figure 3.2 summarises changes in expenditure on education over the period 1996 to Figure 3.2 Education input components from , current prices billions Annual average percentage change Labour Goods and Services Capital Services Total Data for the latest year, 2008, are based on provisional estimates (revised figures will be published in future articles). Education inputs totalled just under 64 billion in 2008, including approximately 4 billion on further education for those aged under 19. In 2008 education was the second largest component of General Government Final Consumption Expenditure (GGFCE), after health; total education input as defined here equates to around 4 per cent of the expenditure measure of Gross Domestic Product The labour and goods and services components are based on: GGFCE estimates for education expenditure in the National Accounts estimates of expenditure on further education for young people (those aged under 19) from national further education skills councils and the devolved administrations The current price capital component is calculated as part of the experimental Volume Index of Capital Services (VICS) developed by ONS. For further information, see Volume of Capital Services: Estimates for 1950 to 2007 (ONS 2009d) Labour costs are the largest component of education expenditure. In 2008, at current prices, labour expenditure was 41 billion, about two-thirds of the total that year. Teachers' pay is the largest element of labour costs. Teacher numbers were relatively stable from 1996 to 2008 but expenditure on support staff, particularly teaching assistants, has increased sharply. This is because of government policies and reforms to reduce unpaid overtime worked by teachers by employing additional support staff. Support staff numbers more than doubled over the same period. Labour costs also include expenditure on indirect support services, such as staff time on policy development, standards setting, finance and training In 2008, 16 billion was spent on goods and services, about one-quarter of all the expenditure on education inputs. This component consists of the goods and services procured from the market sector that are consumed in the production of education services in any given year. This includes items or services such as teaching aids, electricity, building maintenance and transport. Government purchases of ITT, health professional courses and private nursery places are part of the goods and services component The smallest component of education inputs is capital services, estimated at approximately 7 billion in 2008, about a 10th of the total. Goods such as IT equipment and buildings are medium- to longterm investments that can be used for a number of years, and are hence classified as capital items. 10

11 Capital services quantify the flow of inputs from the capital stock into production, through estimates of rental payments. 3.2 Measuring the volume of education inputs In current prices, total expenditure on education inputs increased by just over 120 per cent between 1996 and However, as this includes effects caused by pay and price changes, education inputs are also measured in volume terms. Inputs can be measured directly, for example, using hours worked or a measure of staff numbers in the case of labour inputs. Where data do not allow for a direct measure of inputs, a volume measure can be derived by dividing current price spending figures by an appropriate estimate of price (the deflator) Measuring Productivity, OECD Manual (OECD 2001) recommends that labour inputs should be measured directly. In these estimates, 94 per cent of education labour inputs in the UK are measured in this way. The calculation method uses a breakdown of full-time-equivalent teaching and support staff numbers (teachers are also adjusted for actual hours worked) and weighted together by average salary. The remaining 6 per cent of labour inputs are measured indirectly, by dividing current price expenditure by an appropriate pay deflator. As this small part is a measure of the inputs of central government staff working in education, the most appropriate pay deflator is the public sector Average Earnings Index (AEI) including bonuses This article uses the direct measure of labour described in Labour Inputs in Public Sector Productivity: Methods, Issues and Data (ONS 2009e), which was also used to calculate the previous education estimates presented in ONS (2009a) Goods and services inputs are measured indirectly, using appropriate deflators to remove the effect of price changes. Expenditure on goods and services in the UK can be split into expenditure on both local authorities (around four-fifths) and central government (around one-fifth). Specific price deflators have been derived for each component using ONS producer, retail and service sector price data The capital component is measured using the VICS for local authority and central government education. Further details can be found in ONS (2009d) Figure 3.3 summarises changes in the volume of education inputs by component over the period 1996 to Figure 3.3 Volume of education inputs by component, =100 Annual average percentage change Labour Goods and Services Capital Services Total The volume of UK education inputs grew by 33.3 per cent between 1996 and Goods and services and capital services input volumes grew faster than total volume of inputs, by 78.6 and 48.8 per cent respectively, while the volume of labour, the largest inputs component, grew by 19.0 per cent. 4. Triangulation 4.1 Jointly with parents and guardians and other institutions, schools have responsibilities towards pupils that extend beyond academic outcomes. These wider concerns include physical and mental health, safety and the transfer of social knowledge and skills. Figure 4.1 is reproduced from the supporting Triangulation article (ONS 2009f). It summarises evidence from a range of indicators relating to components of productivity to give context to the estimates presented in this article. 11

12 4.2 This evidence is not necessarily expected to corroborate estimates of productivity change, as sources vary in their scale and scope and therefore rarely cover exactly the same concept. Further discussion on each source of evidence is presented in the supporting Triangulation article to which the section numbers in the table refer. Figure 4.1 Triangulation evidence in summary Indicator [Triangulation article section(s)] Teaching standards [ ] Class sizes, pupil-teacher ratios and pupil-adult ratios [ ] A higher knowledge and understanding of a subject from Further Education [ ] Diet and obesity [ ] Regular sports participation by children [ ] Proportion of 11 to 15-yearolds who regularly smoke [ ] Alcohol and substance abuse [ ] Reported teenage pregnancies [ ] Likely influence of education sector (attribution) Very high measure relates directly to schools High class sizes and ratios can adjust to reflect changes in the school-aged population High measure relates directly education services received Low diet and exercise are only partly influenced by education / school meals High activity in non-school sports likely linked to school sports participation Low social factors likely to play larger influence than schools Low social factors likely to have more influence than schools Very low social factors likely to play much larger influence than schools Key commentary Between 2005/06 and 2007/08, the proportion of primary and secondary schools in which teaching was rated good or above increased in both England and Wales Primary (UK) and secondary (England and Wales) class sizes generally fell between 2000 and 2006; between 1997 and 2009, primary and secondary school pupilteacher & pupil-adult ratios fell in England The proportion of FE students in England who achieved a higher knowledge and understanding of a subject (suitable for higher education or working independently) increased between 1995/96 and 2005/06 The proportion of overweight or obese children in England increased between 1997 and 2007 for both boys and girls The proportion of children in England who regularly participate in sports increased between 2003/04 and 2007/08 The proportion of 11 to 15- year-olds in England who regularly smoke approximately halved between 1982 and Stable proportion of pupils in England drinking between 1998 to 2008; in 2005/06 around a quarter of 15 year olds in GB reported having smoked cannabis in the last 12 months The number of reported teenage pregnancies in England and Wales fell between 1998 and 2007 Agreement with input, quantity or quality changes? Corroborates increases in school quality measures and increase in output Partially corroborates rising quality but evidence on class size and pupil ratios unclear; corroborates rise in inputs through increase in teachers and support staff Further education quality not yet measured but suggests positive link Uncertain. Targeted resources may increase inputs but effects on outcomes not clear Uncertain. Sport has strong impact on health but causality between health and achievement unclear Uncertain. Only small proportion of school population. Targeted resources increase inputs but effects on outcomes unclear Uncertain. Targeted resources may increase inputs but only small part of school population Uncertain. Small positive impact on input through targeted resources but minimal impact on outcomes as only affects very small proportion of pupils Data Source(s) England Office for Standards in Education, Children s Services and Skills Wales Estyn Department for Children, Schools and Families Learning and Skills Council (England) The NHS Information Centre for Health and Social Care Department for Children, Schools and Families The NHS Information Centre for Health and Social Care The NHS Information Centre for Health and Social Care Office for National Statistics 12

13 Indicator [Triangulation article section(s)] Bullying [ ] Proportion of young people not in employment, education or training [ ] Likely influence of education sector (attribution) High schools have key impact on prevalence of bullying Medium/high but dependent not only on level of educational achievement but socio-economic factors Key commentary The proportions of boys and girls aged 11, 13 and 15 years old reporting having been bullied at least twice in the previous two months in 2005/06 broadly fell with age The proportion of year olds in England not in employment, education or training increased between 1996 and 2008 Agreement with input, quantity or quality changes? Uncertain. Positive impact on inputs through targeted resources; no time series evidence to corroborate changes in bullying with changes in output Partial disagreement. Improvements in attainment imply fall in NEETs but there may be distributional effects being missed; also state of the economy has key impact on labour demand Data Source(s) Health Behaviour in School-aged Children: Inequalities in young people s health Department for Children, Schools and Families 5. Next Steps ONS will work with DCSF; the Department for Business, Innovation and Skills (BIS); the devolved administrations and others on a number of developments, summarised below. 5.1 Improvements to the quantity measure Development work will aim to: extend the coverage of the further education quantity measure to cover students of all ages identify publicly funded output from higher education institutions, accounting for the fee and transfer arrangements in each of the devolved administrations and the research work that universities undertake, as well as the number of students they teach 5.2 Improvements to the quality adjustment Development work will aim to: take into account, if possible, research into quantifying attribution and correlation in the wider outcomes relating to children, for example the five outcomes of the Every Child Matters agenda in England explore the possibility of creating a quality adjustment for pre-school education. The National Institute of Economic and Social Research (NIESR) is currently investigating the relationship between outcomes and quality in early years education as part of the Quality Measurement Framework project which is funded by Treasury under the 'Invest to Save Budget'. ONS will be working with NIESR to create a quality adjustment based on the results of this research investigate potential methods for quality-adjusting the quantity of further education output, for example using student retention and achievement data 5.3 Improvements to the measure of the volume of inputs Development work will aim to: improve the inputs measure of further education for those aged under 19, for example by investigating whether expenditure data are available for the three components of further education inputs, which deflators may be appropriate and whether a direct labour measure could be calculated periodically review and improve the inputs measures, as required 13

14 6. References Atkinson T (2005) Atkinson Review: Final Report: Measurement of Government Output and Productivity for the National Accounts, Palgrave Macmillan: Basingstoke. Eurostat (2001) Handbook on Price and Volume Measures in National Accounts, available at: epp.eurostat.ec.europa.eu/portal/page/portal/product_details/publication?p_productcode=ks OECD (2001) Measuring Productivity, OECD Manual, available at: ONS (2007a) Measuring Quality as Part of Public Service Output, available at: ONS (2007b) Methods for Public Service Productivity: Quality Adjusting School Education Output, available at: ONS (2007c) Public Service Productivity: Education, available at: ONS (2009a) Total Public Service Output and Productivity, available at: ONS (2009b) Methods Development: Decomposing Quality Change in the Output of Publicly Funded School Education, available at: ONS (2009c) - Extended Analysis of Output, available at: ONS (2009d) Volume of Capital Services: Estimates for 1950 to 2007, available at: ONS (2009e) Labour Inputs in Public Sector Productivity: Methods, Issues and Data, available at: ONS (2009f) Triangulation available at: 14

15 Annex A: Revisions to Published Estimates A.1 This section reports revisions to ONS estimates of the growth in education productivity since estimates were last published, in Total Public Service Output and Productivity (ONS 2009a). That article presented productivity estimates from 1997 to 2007 while this article covers the period 1996 to Comparisons between the two articles are therefore presented over the longest overlapping period, 1998 to A.2 Estimates are based on multiple administrative data sources which cover uneven periods. Where actual data are not available, forecasts are used (using standard Holt-Winters methods). Revisions to published data may arise when forecasts are replaced by actual data; in addition, they may arise from changes to historical administrative data feeding into the estimates. A.3 Figures A.1 and A.2 summarise the output, inputs and productivity indices presented in both this article and Total Public Service Output and Productivity (ONS 2009a). Figure A.1 Estimates of education output, inputs and productivity as presented in this article, =100 Annual average percentage change Volume of output Volume of inputs Productivity Figure A.2 Estimates of education output, inputs and productivity as previously published (ONS 2009a), = Annual average percentage change Volume of output Volume of inputs Productivity A.4 Figures A.3, A.4 and A.5 summarise the respective revisions to growth in output, inputs and productivity between publications. Figure A.3 Revisions to estimates of education output growth, Percentage points Mean Current Estimate Previous Estimate Revision

16 Figure A.4 Revisions to estimates of education inputs growth, Percentage points Mean Current Estimate Previous Estimate Revision Figure A.5 Revisions to estimates of education productivity growth, Percentage points Mean Current Estimate Previous Estimate Revision A.5 Figure A.6 illustrates revisions to growth in output, inputs and productivity between publications. Figure A.6 Revisions to growth estimates for education output, inputs and productivity, Percentage points Volume of output Volume of inputs Productivity Percentage points Mean Volume of output Volume of inputs Productivity A.6 Over the period 1997 to 2007, the mean revision to output growth is 0.1 percentage points, while the mean revision to inputs growth is -0.1 percentage points. The net result of this is that there has been 16

17 an upward mean revision to productivity growth of 0.1 percentage points. The statistical significance of these revisions has not been investigated since the short span of observations does not allow for hypothesis testing to be conducted. A.7 This article employs a direct labour method, as described in Labour Inputs in Public Sector Productivity: Methods, Issues and Data (ONS 2009e), while an indirect labour measure was used within the overall inputs calculations in ONS (2007c). Improvements to the goods and services method include changes to the underlying price indices and a slight change in methodology to better account for expenditure patterns. The capital component is measured using the VICS for local authority and central government education. Further details can be found in Volume of Capital Services: Estimates for 1950 to 2007 (ONS 2009d). In addition, expenditure data that are deflated and used to aggregate volume indices are also subject to revisions. 17

18 Annex B: Areas for Improvement in the Schools Quality Adjustment Figure B.1 Areas for improvement in the schools quality adjustment Problem Direction of Error Proposed/Completed Developments The model of education has a major impact on the quality measure: any chosen model likely to oversimplify true model Dependent on the model used and how close it is to the true model Exam attainment of only one cohort of pupils in education in any given year is used to approximate quality changes across the whole pupil body in that year Time lag: method is slow to pick up improvements in early years of schooling Change in examination standards Narrow focus: examination attainment is an important school outcome but not the only one If improvements are made in primary education, these will not be picked up and the current method would understate the quality gain If primary education improves, measure understates quality gain If examinations standards are not consistent over time, then current measure may overstate or understate quality change If contribution of schools to wider outcomes is not improving at same rate as examination grades, then measure may overstate quality gain See Methods Development: Decomposing Quality Change in the Output of Publicly Funded School Education (ONS 2009b) See ONS (2009b) See ONS (2009b) See Public Service Productivity: Education (ONS 2007c) Awaiting the results of research into the relationship between outcomes and quality in early years education 18

19 A National Statistics publication National Statistics are produced to high professional standards set out in the Code of Practice for Official Statistics. They are produced free from political influence. About us The Office for National Statistics The Office for National Statistics (ONS) is the executive office of the UK Statistics Authority, a non-ministerial department which reports directly to Parliament. ONS is the UK government s single largest statistical producer. It compiles information about the UK s society and economy, and provides the evidence-base for policy and decision-making, the allocation of resources, and public accountability. The Director-General of ONS reports directly to the National Statistician who is the Authority's Chief Executive and the Head of the Government Statistical Service. Published with the permission of the Office of Public Sector Information (OPSI) You may use this publication (excluding logos) free of charge in any format for research, private study or internal circulation within an organisation providing it is used accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. For re-use of this material you must apply for a Click-Use Public Sector Information (PSI) Licence from: Office of Public Sector Information, Crown Copyright Licensing and Public Sector Information, Kew, Richmond, Surrey, TW9 4DU Tel: The Government Statistical Service The Government Statistical Service (GSS) is a network of professional statisticians and their staff operating both within the Office for National Statistics and across more than 30 other government departments and agencies. Contacts This publication For information about the content of this publication, contact Richard Wild Tel: Richard.Wild@ons.gsi.gov.uk Other customer enquiries ONS Customer Contact Centre Tel: International: +44 (0) Minicom: info@statistics.gsi.gov.uk Fax: Post: Room 1015, Government Buildings, Cardiff Road, Newport, South Wales NP10 8XG Media enquiries Tel: press.office@ons.gsi.gov.uk Copyright and reproduction Crown copyright