Nationally reported data for each local authority reveals that spend on children’s services per child varies significantly between authorities, from £292 to £1,254 (based on revenue outturn returns for FY16/17 and the local population aged 0-25). Standardising for similar outcomes, spend per child still varies from £299 to £805. This work, commissioned by the Local Government Association (LGA) and undertaken by Newton, seeks to shed light on the factors driving variation of spend on children’s services, establishing why some authorities spend significantly more than others, whilst achieving similar levels of outcome.The work comprises two major projects. The first involves a review and analysis of national demographic, geographic and economic data for all councils in the country, to identify correlations with children’s services spend. The second project, a detailed, on-the-ground study of the factors influencing spend which could not be quantified through national data alone, was completed with a sample of eight councils. The factors identified included analysis of financial reporting and the influence of council and partner organisation practice. Through correlation analysis of 18 demographic, economic and geographic measures, it has been possible to identify a set of five factors that explain approximately half of the variation in spend seen nationally across all authorities. These factors are largely outside of the control of councils, and certainly sit outside the control of children’s services. As a consequence, variation in what authorities spend on children’s services (per head of child population) is inevitable. It is not logical to expect authorities to converge on a single ‘right’ value of spend. Of the five drivers of variation, the IDACI measure of deprivation was found to be the single most significant factor. Alone, it explains 31% of variation. When combined with the other four factors (size of population aged 0-25, amount of disposable household income, levels of unemployment and levels of crime) the resulting model describes just over 50% of spend variation. Analysis of reported spend, according to the revenue outturn financial returns, echoes previous research that has shown variation in the accuracy of how these returns are compiled. This makes it difficult to draw a like-for-like comparison of spend between councils. The treatment of grants, coding of spend to areas of the revenue outturn and allocation of central overheads were the three key areas of inaccuracy identified. The complexity is compounded by the fact that the guidelines for the returns do not permit a like-for-like comparison of spend between authorities, specifically around how spend on asylum-seeking children is set out. In the sample of authorities with whom in-depth analysis of the financial returns was completed, the overall contribution of this factor to spend variation was small. However, this is not necessarily representative of the broader national picture since examples were seen where individual council spend changed by up to 12%. Nationally, spend on looked after children and safeguarding children and families (which covers core spend on statutory child in need and child protection social work) are both the largest areas of spend and also the areas where spend varies most between authorities. Study of these areas across the sample authorities identified that taking the right local steps to achieve the best possible outcomes for children, young people and families could reduce spend variation between the sample group of councils by 13%. The changes required to achieve this consistently, whilst more in the control of councils than other local partners, are nonetheless complex, potentially costly and challenging to achieve, with no ‘one size fits all’ solution across authorities. Opportunities exist to explore the role of partnership working and practice consistency further to assist authorities in developing local improvement plans. Spend on other areas of children’s services outside of looked after children and safeguarding children and families makes up 30% of spend on average but presented a source of variation between the sample of authorities. In one authority the figure was 22%, whilst in another it was 56%. Within this, different approaches to early help and preventative services were identified. The methods employed in this study were not designed to identify a correlation between spend on early help / preventative services and lower spend at higher tiers. It was noticeable that councils spending more on these services tended to do so primarily as a result of political or strategic commitment to the importance of providing help early. Greater engagement in the building of the evidence-base around early help and preventative services, at a time when such spend is coming under pressure, is likely to be important if funding for them is to be maintained. By taking two different approaches this work has described some of the drivers behind up to 50% of the variation in spend on children’s services. More importantly perhaps, the work also re-emphasises that we should expect to see some degree of variation driven by wider economic and geographical factors; our modelling suggests that a range from £334-£883 could be expected as the result of five factors alone. This coupled with variation in financial reporting practices means that caution needs to be taken when comparing one authority with another. Rather; the narrative around spend needs to be based on local system understanding and the appropriate cost of achieving the ideal outcomes for children and families. The scope of this work, looking specifically at factors driving variation in spend on children’s services between councils, is one of many questions and challenges facing the sector. Further work in this area is required to address the next logical questions around factors driving increases in demand and the future funding arrangements for children’s services, including the total quantum of spend.