Downloaded from adc.bmj.com on March 22, 2010 - Published by group.bmj.com
Leading article
The early evaluation protocol to 2005 has Evaluation and value of Sure Start been summarised, 9 and detailed reports are available at http://www.ness.bbk. ac.uk, with some fi ndings summarised E Melhuish, J Belsky, J Barnes here. Government decisions effectively ruled out a randomised controlled trial; A decade ago, the Cross-Departmental the corresponding fi ndings, this report hence, a quasi-experimental design was Review of Services for Young Children provides an overview of the evolution and used with its consequent limitations to concluded that disadvantage among impact of the Sure Start programme. As compare Sure Start populations with young children was increasing and it was individuals involved in the evaluation of equivalent populations not receiving Sure more likely that poor outcomes could be the overall initiative, we attempt to pro- Start. An independent review of early prevented when early intervention was vide a summary of what has been learnt fi ndings of the evaluation work and a cri- 1 undertaken. The review also noted that thus far. Community control was exer- tique of the methodology is available. 10 current services were uncoordinated and cised through local partnerships, compris- patchy, and recommended that there ing everyone concerned with children, COMMUNITY-LEVEL CHANGE should be a change in service design and including health, social services, educa- delivery. It suggested that programmes tion, private and voluntary sectors, and Sure Start was based on the premise should be jointly planned by all relevant parents. Funding fl owed from central gov- that children and families could be bodies and be area based, with all chil- ernment directly to programmes, which affected by the programme directly, via dren under 4 years old and their families were independent of local government, services, and indirectly, via community in an area being clients. In July 1998, although local departments of educa- changes engendered by the programme. the then Chancellor of the Exchequer, tion, social services, etc, and health trusts Community characteristics were tracked Gordon Brown, introduced Sure Start, would typically be part of the partnership. over 5 years. Some improvements were which is aimed at providing quality ser- All programmes were expected to provide detected but could not be causally 11 12 vices for children under 4 years old and (1) outreach and home visiting; (2) sup- linked to Sure Start. Overall, Sure their parents. 2 The original intent of the port for families and parents; (3) support Start areas became home to more young programme design was to focus on the for good quality play, learning and child- children while the proportion in house- 20% most deprived areas, which included care experiences for children; (4) primary holds dependent on benefi ts decreased around 51% of children in families with and community healthcare and advice markedly. For instance, the proportion incomes 60% or less than the national about child health and development and of children under 4 years in “workless” median (offi cial poverty line). 3 family health; and (5) support for people households in Sure Start areas dipped Over 3 years, £542 million was made with special needs, including help getting below 40%, having started at 45% available, with £452 million desig- access to specialised services but without in 2000–2001. Some aspects of crime nated for England. In England, 250 pro- specifi c guidance as to how. 4 While evi- and disorder also improved, notably grammes were planned by 2001–2002 to dence derived from programmes with burglary, school exclusions and unau- support 187 000 children, 18% of poor clear unambiguous protocols for services thorised school absences. Moreover, children under 4 years old. Typically, were used to justify Sure Start, Sure Start children 11 years and older demonstrated a programme was to include 800 chil- programmes did not have a prescribed improved academic achievement. While dren under 4 years old, with £1250 per “protocol” and had freedom to improve infant health did not improve, reductions annum per child at the peak of fund- and create services as they wished. 5 This in emergency hospitalisations of children ing. Programmes were to run for at least was in contrast to examples of interven- (0–3 years) for severe injury and for lower 10 years with funding peaking at year tions with clear models of provision and respiratory infection suggested improved 3 and declining from year 6 to none at demonstrable effectiveness used to justify healthcare. Also, the percentage of chil- year 10. Some funding would fall to local Sure Start. 6–8 dren identifi ed with special educational authorities, made available by “reshap- The speed of funding was often over- needs or eligible for disability benefi ts ing” services. This investment utterly whelming in a sector previously starved increased over 2000–2005, suggesting transformed early-year services while of support. Only 6% of the 1999 alloca- improved health screening. representing a relatively small contribu- tion was spent in that year. Despite this tion from the perspective of treasury— slow start, and without any informa- EARLY EFFECTS ON CHILDREN/ just 0.05% of public expenditure. tion on the success of the initiative, the FAMILIES In this brief summary, we cover the treasury expanded Sure Start from 250 A cross-sectional impact study of children history and impact of this ambitious by 2002 to more than 500 programmes and families in Sure Start and non-Sure initiative. While more than 40 reports by 2004, thereby more than doubling Start (Sure Start-to-be) areas provided and peer-reviewed publications have expenditure to almost £500 million by mixed news. 13 14 Although there were documented the different phases of the 2003–2004. Thus, Sure Start became a some main effects for Sure Start on fam- National Evaluation of Sure Start with cornerstone of the campaign to reduce ily and child outcomes, some results var- child poverty. ied by subgroup. Specifi cally, 3-year-old The National Evaluation of Sure Start children of non-teen mothers (86% of Institute for the Study of Children, Families and Social (NESS) was commissioned in 2001. sample) in communities receiving Sure Issues, Birkbeck, University of London, London, UK Programme diversity posed challenges in Start exhibited positive effects associ- Correspondence to: Edward Melhuish, Institute that there were not several hundred pro- ated with Sure Start programmes in for the Study of Children, Families and Social Issues, grammes delivering one intervention, but terms of fewer behaviour problems and Birkbeck, University of London, 7 Bedford Square, several hundred unique interventions. greater social competence as compared London WC1B 3RA, UK; [email protected]
Arch Dis Child 2010; 95 :159–161. doi:10.1136/adc.2009.161018 159 Downloaded from adc.bmj.com on March 22, 2010 - Published by group.bmj.com
Leading article with those in comparison communities, of services and guidelines, and were con- programmes throughout the entire lives and evidence indicated that these effects trolled by local authorities rather than of the children. Also, programmes had for children were mediated by Sure Start the central government. the opportunity to learn from the ear- effects of less negative parenting for lier phase of the evaluation, especially non-teens. On the other hand, adverse LATEST EFFECTS ON CHILDREN AND with respect to a greater effort to reach effects emerged for children of teen FAMILIES the most vulnerable households. Thus, mothers (14% of sample) in Sure Start In the longitudinal investigation of chil- differences in the amount of exposure to areas in terms of lower verbal ability and dren and families seen at 9 months and programmes and the quality of Sure Start social competence and higher behaviour 3 years of age, comparisons of children programmes may well account for the problems. Also, children from workless and families in Sure Start areas with initial adverse effects for the most disad- households (40% of sample) and from those in similar non-Sure Start areas vantaged children and families and the lone-parent families (33% of sample) in revealed benefi cial effects for children subsequent benefi cial effects for almost Sure Start areas scored lower on verbal and families living in Sure Start areas, all children and families living in Sure ability than equivalent children in com- when children were 3 years old on 7 of Start areas. It is also noteworthy that a parison communities. the 14 outcomes assessed. 23 24 Children randomised controlled trial in Sure Start in Sure Start areas showed better social programmes in Wales has found that the VARIABILITY IN PROGRAMME development, exhibiting more positive Incredible Years parenting programme EFFECTIVENESS social behaviour and greater indepen- can improve behavioural and emotional NESS also examined why some pro- dence/self-regulation than their counter- outcomes for preschool children consid- grammes might be more effective than parts in non-Sure Start areas. This result ered to be at risk of developing conduct 25 others. Detailed information over sev- was partially a consequence of parents disorder. eral years on 150 programmes was sys- in Sure Start areas manifesting less neg- Sure Start has been evolving, and ongo- tematically rated on 18 dimensions of ative parenting, as well as a better home ing research has partly infl uenced this implementation refl ecting adherence to learning environment. Also, families process. Later developments have con- Sure Start principles. 15–18 Programmes in Sure Start areas reported using more siderably clarifi ed guidelines and service that were rated high on realising one child and family-related services than delivery. It is plausible that the improved principle tend to score high on others, families in non-Sure Start areas. Two results in the evaluation of Sure Start and better implemented programmes additional fi ndings—that children in refl ect actual changes in the impact of appeared to yield greater benefi ts. 19 Sure Start areas received more recom- Sure Start programmes resulting from 20 While the evidence was not over- mended immunisations and had less the increasing quality of services, greater whelming, it was consistent with the accidental injuries than those in other attention to the hard to reach, the move theory about which programmes should areas—could have been an artefact of to children’s centres, as well as the prove most effective and provided guid- the 2-year difference in the date of data greater exposure to the programme of ance as to what works for Sure Start gathering for the two groups, as these children and families in the latest phase Programmes. Also, some evidence indi- two outcomes tended to improve over of the impact evaluation. The results are cated that health-led programmes had time nationally. modest but suggest that the value of Sure some advantages, possibly refl ecting Start programmes is improving. The their better access to birth records and identifi cation of the factors associated health visitors providing a ready-made CONCLUSIONS with more effective programmes has home-visiting service generally accepted The latest fi ndings differ markedly from propelled recent improvements in Sure by disadvantaged families. earlier fi ndings. Whereas earlier, the Start Children’s Centres and may be part most disadvantaged 3-year-old children of the reason for the improved outcomes and their families (ie, teen parents, lone for children and families now found for CHANGES TO SURE START parents and workless households) were Sure Start. As the initial NESS fi ndings indicated doing less well in Sure Start areas, while that Sure Start programmes were not somewhat less disadvantaged children Funding The UK Department of Children Schools and Families funded the research that is the basis of this having the intended impact and evidence and families benefi ted (ie, non-teen par- article. from another project, Effective Provision ents, dual parent families and working of Pre-school Education, 21 22 showed that households), the latest evidence indicates Competing interests None. integrated Children’s Centres were par- benefi ts for all sections of the popula- Ethics approval This study was conducted with ticularly benefi cial to children’s devel- tion served. Various explanations can be the approval of the South West Multicentre Research opment, the government decided to offered for the differences between the Ethics Committee. transform Sure Start programmes into 2005 and 2008 fi ndings. Although it is Children’s Centres. This was announced not possible to entirely eliminate meth- Provenance and peer review Commissioned; exter- in 2005 alongside a transfer of the new odological explanations, it seems likely nally peer reviewed. Sure Start Children’s Centres to local that the contrasting results accurately Accepted 9 October 2009 authority control. This politically refl ect the contrasting experiences of inspired move ensured that Sure Start children and families in Sure Start areas REFERENCES Children’s Centres became embedded in the two phases. Whereas the 3-year- 1. HM Treasury. Modern public services for Britain: within the welfare state by statute, mak- olds in the fi rst phase were exposed to investing in reform. Comprehensive spending ing it diffi cult for any future government “immature” programmes—and probably review: new public spending plans. London: HMSO, 1998. to eradicate. Hence, post 2005, Sure not for their entire lives—the 3-year- 2. HM Treasury. Comprehensive spending review: Start programmes became Children’s olds and their families in the second cross departmental review of provision for young Centres with a more clearly specifi ed set phase were exposed to better developed children. London: HMSO, 1998.
160 Arch Dis Child 2010; 95 :159–161. doi:10.1136/adc.2009.161018 Downloaded from adc.bmj.com on March 22, 2010 - Published by group.bmj.com
Leading article
3. Melhuish E, Hall D. The policy background to Sure area-based early intervention work? Bristol: Policy 19. NESS Research Team . Variation in Sure Start. In: Belsky J, Barnes J, Melhuish E, eds. The Press, 2007:173–94. Start Local Programmes effectiveness: early national evaluation of Sure Start: does area-based 12. Barnes J, Cheng H, Howden B, et al . Changes in preliminary findings . Surestart Report 14. early intervention work? Bristol: Policy Press, the characteristics of SSLP areas between 2000/01 London: DfES, 2005. http://www.ness.bbk. 2007:3–21. and 2004/05 . London: DCSF, 2007. ac.uk/impact/documents/1184.pdf (accessed 21 4. Sure Start Unit. Sure Start: guide for trailblazer 13. NESS Research Team. Early impacts of Sure Jul 2009). programmes. London: Department for Education Start local programmes on children and families . 20. Melhuish E, Belsky J, Anning A, et al ; NESS and Employment, 1998. Surestart Report 13. London: DFES 2005. http:// Research Team. Variation in community 5. Glass N. Sure Start: the development of an www.ness.bbk.ac.uk/impact/documents/1183.pdf intervention programmes and consequences for early intervention programme for young children (accessed 21 Jul 2009). children and families: the example of Sure Start in the United Kingdom. Children & Society 14. Belsky J, Melhuish E, Barnes J, et al. Effects Local Programmes. J Child Psychol Psychiatry 1999; 13 :257–64. of Sure Start Local Programmes on children and 2007; 48 :543–51. 6. Ramey CT , Campbell FA, Burchinal M, et al. families: early fi ndings from a quasi-experimental, 21. Sylva K , Melhuish E, Sammons P, et al . Effective Persistent effects of early childhood education on cross sectional study. BMJ 2006; 332 :1476–8 pre-school provision . London: Institute of Education, high-risk children and their mothers. Appl Dev Sci 15. Anning A, Ball M. Living with Sure Start: human 2004. 2000; 4:2–14. experience of an early intervention programme. 22. Melhuish EC , Sylva K, Sammons P, et al. The 7. Reynolds AJ , Temple JA, Robertson DL, et al. In: Belsky J, Barnes J, Melhuish E, eds. The early years. Preschool infl uences on mathematics Long-term effects of an early childhood intervention national evaluation of Sure Start: does area-based achievement. Science 2008; 321 :1161–2. on educational achievement and juvenile arrest: a early intervention work? Bristol: Policy Press, 23. NESS Research Team. The impact of Sure Start 15-year follow-up of low-income children in public 2007:97–112. Local Programmes on three year olds and their schools. JAMA 2001; 285 :2339–46. 16. Anning A, Ball M. Improving services for young families . Surestart Report 27. London: DCSF, 2008. 8. Schweinhart LJ , Barnes H, Weikhart, D. children: from Sure Start to children’s centres. www.surestart.gov.uk/ _ doc/P0002519.pdf Signifi cant benefi ts: the High/Scope Perry Pre- London: Sage, 2007. (accessed 21 Jul 2009). School Study through age 27. Michigan: High/Scope 17. Meadows P. The cost and benefi ts of Sure Start 24. Melhuish E, Belsky J, Leyland AH, et al ; National Press, 1993. Local Programmes. In: Belsky J, Barnes J, Melhuish Evaluation of Sure Start Research Team. Effects 9. Belsky J , Barnes J, Melhuish E, eds. The National E, eds. The national evaluation of Sure Start: does of fully-established Sure Start Local Programmes Evaluation of Sure Start: does area-based early area-based early intervention work? Bristol: Policy on 3-year-old children and their families living in intervention work? Bristol: Policy Press, 2007. Press, 2007:113–30. England: a quasi-experimental observational study. 10. Rutter , M. Is Sure Start an effective preventive 18. Tunstill J, Allnock D. Sure Start Local Lancet 2008; 372 :1641–7. intervention? Child Adolesc Ment Health Programmes: an overview of the implementation 25. Hutchings J, Bywater, T, Daley D, et al. 2006; 11 :135–41. task. In: Belsky J, Barnes J, Melhuish E, eds. Parenting intervention in Sure Start services for 11. Barnes J. How Sure Start Local Programmes The national evaluation of Sure Start: does area- children at risk of developing conduct disorder: areas changed. In: Belsky J, Barnes J, Melhuish based intervention work? Bristol: Policy Press, pragmatic randomised controlled trial. BMJ E, eds. The national evaluation of Sure Start: does 2007:79–96. 2007; 334 :678.
Arch Dis Child 2010; 95 :159–161. doi:10.1136/adc.2009.161018 161