Technical report: Key Stage 4 Outcomes for different groupings of disadvantaged and vulnerable children

Introduction

The Children’s Commissioner’s Office’s programme of data work on childhood vulnerability aims to shine a light on the children growing up with additional needs or issues that may need more help to succeed. These include educational disadvantage, severe poverty, special educational needs or disabilities, risks of abuse and neglect, mental health needs, youth violence and exploitation, homelessness, and other issues. Our work on this area has set out to show the scale and nature of childhood vulnerability and how they interact.

Many children grow up facing at least one vulnerability: even before the pandemic – 2.3 million children in England were growing up in a family vulnerable family with a serious and complex need¹. Even more concerning are the children who face multiple overlapping vulnerabilities – risks that compound each other and collectively chip away at children’s ability to thrive. Unfortunately, it is these kinds of more severe vulnerabilities where data and insights are lacking the most.

Our work on vulnerability has made valuable inroads, by showing the 100,000 children living in a household affected by all of the so called ‘toxic trio’ issues – domestic abuse, parental mental health issues and parental drug and alcohol dependency² – or the 140,000 teenagers with two or more high needs (including dropping out of school, special educational needs, or having a social worker)³. Despite this, it is still too difficult to assess the numbers, experiences and outcomes of the children that we should be most worried – those living with multiple and persistent vulnerabilities.

One area where we can make more progress is in understanding the impact of vulnerability on education. It is well known that disadvantaged children tend, on average, to have lower levels of educational attainment. Children eligible for free school meals (FSM) are only around half likely as likely as the rest of the cohort to achieve a strong pass in English and Maths GCSE. Furthermore, children with identified special educational needs (SEN) are only a quarter as likely to achieve this benchmark, compared to children without SEN.⁴. The Department for Education’s Child in Need (CIN) review⁵ indicates that children who have ever needed a social worker are around a half as likely to achieve this, compared to other children. More recent evidence from the Bristol and Oxford Universities shows that within the group of children too often we consider one vulnerability at a particular point in time. These separate pieces of evidence are powerful and important on their own, but what has too often been missed is that in many cases they could be examining the same child from three different lenses. In other words, these groups often overlap – and most vulnerable children will belong to more than one group.

Data analysed for this report, focusing on the cohort of children taking GCSE exams in 2019, shows the degree of overlap. Of the children who have needed a social worker in the past six years, 4 in 5 also had another disadvantage (FSM or SEN) in the past six years. Of the children who had been FSM in the past six years, more than half also been CIN or had SEN over that period as well. And nearly half of the children who had had SEN over the past six years and had also been CIN or FSM over that period. As these disadvantages compound, so too does the additional support that may needed. Some children – around 1 in 25 of the Year 11 cohort – had all three of these characteristics.

This report also shows the extent to which these issues persist over time. It finds that less than half (44%) of the pupils who have had a social worker have only done so for one year; the rest have had a social worker for multiple years. Among those who have had SEN, only around three quarters have done so for more than a year, and among those who have had FSM this is more than four-fifths.

As well as looking at the intersections of these different groups, this analysis also looks within these groups to see how education outcomes vary amongst children with needs identified at different thresholds – for example, children in care or children with an Education, Care and Health Plan (EHCP).

This report finds only around 2 in 5 of the broader group of children – those who have been CIN, FSM or SEN – achieved a pass in GCSE English and Maths in 2019. But children who only have one vulnerability will still do well on average. For example, nearly two thirds of children who only had a social worker – and not FSM or SEN – achieved these passes. The same is true of children who only had FSM. These rates are actually in line with the average pass rate across all pupils.

But as vulnerabilities coincide and become more severe or persistent, outcomes can rapidly deteriorate. Of the children who have had a social worker and been on FSM and had SEN, only 13% passed GCSE English and Maths. The rate falls to as low as 4% if the child was continually on FSM throughout school and also had an EHCP.

The proportions of children who have these various types and levels of need is therefore a crucial way of understanding the overall of educational vulnerability. Overall, these groups collectively account for the clear majority of children failing to achieve basic qualifications in England. Of all the children not achieving levels 9-4 in GCSE English and Maths, nearly three-quarters – 72% – have been CIN, FSM or SEN in the last six years.

We also need to recognise that these rates vary considerably across the country: in some local areas, nearly 3 in 4 pupils have been CIN, FSM or SEN in the past six years. But there is also wide variation in the outcomes that these children achieve: in some areas more half of these pupils managed to pass English and Maths GCSE, whereas in other areas less than a third of this group managed to do so.

With the data available, it is difficult to explain what is driving this local variation, even after taking into account other local factors including deprivation and rates of access for SEN and CIN support. We still need to know more about what it is that enables some of these highly vulnerable children to succeed – either because of the area they live in, or because of personal or family factors that enable them to thrive regardless. What is clear, however, is that re-examining the data on disadvantaged children in this way helps us to see which groups – on average – are the most left-behind and need the most help

Executive summary

This analysis examines the Key Stage 4 (KS4) results for the following groupings of disadvantaged and vulnerable children:

• Those who have had an open episode with children’s services in the last 6 years (CIN 6)
• Those who have been eligible for free school meals in the last 6 years (FSM 6)
• Those who have had any identified special educational need in the previous 6 years (SEN 6).

We examine Key Stage 4 results for children at the intersections of these groups to examine what (if any) additional disadvantage children face when they have multiple identified vulnerabilities. We also look within these groups and examine how KS4 results vary amongst children with needs identified at higher thresholds. We investigate this amongst the most recent cohort of children to take GCSEs that were not affected by the COVID-19 pandemic - those who sat their GCSEs in summer 2019. The main outcome measure in this analysis is whether these children achieved Levels 9-4 in English and Maths GCSE.

Our key questions to answer are:

• What are the Key Stage 4 results for children with varying combinations of these CIN 6, FMS 6 and SEN 6 characteristics?

• How do results vary in relation to the length of time that they have been CIN/FSM eligible/identified SEN over the past 6 years?

• How do these results vary by local authority for children with combinations of these characteristics?

Overall this analysis echoes previous work suggesting that large numbers of children have these identified vulnerabilities and that (on average) they have notably worse outcomes at Key Stage 4. However, the key addition is that it also demonstrates there is considerable variation within these groups in children’s Key Stage 4 results based on their combinations of these vulnerabilities.

• Children with any of these CIN/FSM/SEN 6 – that is, CIN 6 or FSM 6 or SEN 6 – characteristics represent a substantial proportion of this cohort, accounting for just under 1 in 2 children sitting GCSE exams in 2019.

• They have notably lower rates of achieving levels 9-4 in KS4 English and Maths. Around 4 in 10 of this group achieve at least a level 4 in these subjects compared to a cohort average of just over 6 in 10. Around 1 in 4 achieve at least a level 5, compared to 4 in 10 in the cohort overall.

• Children with any of these CIN/FSM/SEN 6 characteristics account for 72% of the children in this cohort not achieving levels 9-4 in KS4 English and maths and 61% of those not achieving levels 9-5.

• Rates of achieving at least level 4 in these subjects are lowest among those that have had an EHC plan in the last 6 years (12%), and among those that have been looked after at some point in the last 6 years (23%).

• The specific combination of children’s CIN/FSM/SEN 6 status makes substantial differences to their performance at KS4: Nearly two thirds of children that are CIN 6 only – i.e. CIN 6 but not FSM 6 and not SEN 6 – achieve at least a level 4 in English and Maths. This is similar to the average rate for the cohort as a whole (64%). The lowest rates are amongst children with an EHC plan in the last 6 years in combination with an open CIN episode and being FSM eligible throughout the 6 years achieving at least a level 4 in English and Maths (4%).

• Children with an EHC plan in the last 6 years combination with other CIN/SEN/FSM 6 characteristics have the lowest rates of achieving at least a level 4 in English and Maths.

• There is substantial variation by local authority in rates of these children achieving at least a level 4 in English and Maths.

  • Rates of FSM 6 children achieving this range from 28% to 68% across local authorities.
  • Rates of CIN 6 children achieving this range from 20% to 54%.
  • Rates for children with SEN support but no EHC plan in the last 6 years range from 17% to 56%.

Definition of cohort

We base our analysis on the cohort of children who took Key Stage 4 (KS4) exams in 2019 and can be matched in the Summer term school census 2019 (n = 548,610). Note that since the School Census primarily covers state-funded schools only, the resulting matched cohort largely excludes who were not enrolled in a state school⁶.

We then link this base cohort to every termly pupil level school census over the six year period from 2013/14 to 2018/19 (via the anonymised pupil matching reference) in order to add information on each pupil’s SEN and FSM status in each term during this period⁷. We also link in the children in need (CIN) census datasets over the same time period to add information on whether each pupil had an open CIN episode with children’s social care.

We then link this base cohort to the termly pupil level school census (hereafter the NPD) and annual children in need census datasets from 2014-2019 (inclusive) via their anonymised pupil matching reference and UPN respectively⁸. This creates a longitudinal record of this cohort’s number of terms with identified SEN and FSM eligibility as well as years when they have an open CIN episode with children’s services.

Rates of disadvantage

The main disadvantage indicators in this analysis are:

• Whether a pupil had an open episode that has been assessed as ‘in need’ with children’s services for any length of time in the last 6 years (CIN 6)
• Whether a pupil had been eligible for free school meals in the last 6 years (FSM 6)
• Whether a pupil had any identified special educational need in the previous 6 years (SEN 6).

Table 1 and Figure 1 below show the proportions of the cohort that were CIN/SEN/FSM in the last 6 years. Just under 1 in 2 children in this cohort had any of these characteristics in the last 6 years and around 1 in 20 had all 3 at some point.

Table 1: Proportions of cohort that are CIN 6/SEN 6 or FSM 6

Disadvantage type	Disadvantage indicator	Number of pupils	% of cohort
CIN	Any CIN in last 6 years	68138	12
FSM	Any FSM in last 6 years	134207	24
SEN	Any SEN in last 6 years	160444	29
	Any SEN without a statement/EHC plan in the last 6 years	144938	26
	Any statement/EHC plan in the last 6 years	22581	4
Summary	Any CIN/SEN/FSM in last 6 years	248298	45
	CIN 6 & FSM 6 & SEN 6 in last 6 years	23960	4
	Any CIN/SEN/FSM in last 6 years exc. EHC plans	242710	44
	CIN 6 & FSM 6 & SEN 6 in last 6 years exc. EHC plans	20257	4

Figure 1: Venn diagram of intersections between CIN 6, SEN 6 and FSM 6 groups

Figure 1 also shows a notable degree of overlap between children in these groups. For example, 79% of the pupils in the CIN 6 group are also a member of the SEN 6 or FSM 6 groups. More than half – 57% – of the pupils in the FSM 6 group are also in the SEN 6 or CIN 6 groups. And nearly half of the pupils in the SEN 6 group are also in the FSM 6 or CIN 6 groups.

Tables 2, 3, 4 and demonstrates that these overlaps are larger than would be expected if the FSM 6, SEN 6 and CIN 6 were distributed randomly. Table 2 demonstrates CIN 6 children are over twice as likely to have had any identified SEN in the past 6 years compared to non-CIN 6 children.

Table 2: Proportions of CIN 6 group that are also SEN 6

SEN 6	CIN 6	Not CIN 6
Not SEN 6	45% (30,352)	74% (357,814)
SEN 6	55% (37,786)	26% (122,658)

Table 3 demonstrates a similar strong association between children being CIN 6 and FSM 6. CIN 6 children are nearly three times as likely to have also been eligible for free school meals in the past 6 years compared to non-CIN 6 children.

Table 3: Proportions of CIN 6 group that are also FSM 6

FSM 6	CIN 6	Not CIN 6
FSM 6	59% (39,900)	20% (94,307)
Not FSM 6	41% (28,238)	80% (386,165)

Table 4 demonstrates a similar strong association between children being SEN 6 and FSM 6. CIN 6 children are nearly twice as likely to have also been eligible for free school meals in the past 6 years compared to children with no identified SEN.

Table 4: Proportions of SEN 6 group that are also FSM 6

FSM 6	Not SEN 6	SEN 6
FSM 6	19% (73,442)	38% (60,765)
Not FSM 6	81% (314,724)	62% (99,679)

Table 5 and Figure 2 demonstrates the proportions of the cohort with identified higher threshold needs. Just over 1% of the cohort have been looked after in the last 6 years (LAC 6). Around 1 in 10 children have been either looked after, had an EHC plan or been FSM eligible for all of the possible terms in the last 6 years.

Table 5: Proportions of cohort with identified higher threshold needs in the last 6 years

Characteristic	Number of pupils	% of cohort
EHC plan in last 6 years	22581	4.1
LAC in last 6 years	7540	1.4
FSM for all possible terms in last 6 years - Persistent FSM	38848	7.1
Any EHC/LAC/Persistent FSM in last 6 years	62020	11.3
CIN 6 & FSM 6 & SEN 6 in last 6 years	214	0.0

Figure 2: Venn diagram of intersections between LAC 6, EHC 6 and Persistent FSM groups

Rates of persistent disadvantage

The sections below additional detail on the length of time for which these pupils had these disadvantage indicators – SEN, FSM or CIN – in the 6 years prior to their Key Stage 4 exams.

Findings: Key Stage 4 outcomes for CIN/SEN/FSM 6 children

Table 22 demonstrates that children in the CIN 6, FSM 6 and SEN 6 groups are considerably less likely than the overall KS4 population to achieve levels 9-4 in KS4 English and Maths. 42% of those with any CIN/FSM or SEN in the last 6 years achieved this benchmark (43% excluding children with an EHC plan in the last 6), compared with a cohort average of 64%. Lowest rates are amongst children with any EHC plans over the last 6 years (12%) and children that were looked after over the last 6 years (23%)¹⁰.

Table 22: Proportions of each group achieving levels 9-4 in KS4 english and maths

Disadvantage type	Disadvantage indicator	% achieving 9-4 KS4 English and Maths	% achieving 9-5 KS4 English and Maths
CIN	Any CIN in last 6 years	34% (22,976)	17% (11,921)
	Any Child Protection Plan in the last 6 years	27% (3,331)	13% (1,624)
	Children looked after at any point in the last 6 years	23% (1,743)	10% (776)
	Children persistently CIN over 6 years	20% (2,569)	9% (1,145)
FSM	Any FSM in last 6 years	43% (57,578)	24% (31,696)
	Children persistently FSM eligible over 6 years	37% (14,548)	20% (7,720)
SEN	Any SEN in last 6 years	32% (50,935)	16% (24,920)
	Any SEN without a statement/EHC plan in the last 6 years	34% (49,201)	17% (24,033)
	Any statement/EHC plan in the last 6 years	12% (2,708)	6% (1,375)
	Children persistently SEN over 6 years	17% (7,381)	8% (3,345)
Summary	Any CIN/SEN/FSM in last 6 years	42% (105,137)	23% (56,790)
	CIN 6 & FSM 6 & SEN 6 in last 6 years	13% (3,199)	5% (1,242)
	Any CIN/SEN/FSM in last 6 years exc. EHC plans	43% (104,065)	23% (56,208)
	CIN 6 & FSM 6 & SEN 6 in last 6 years exc. EHC plans	15% (3,047)	6% (1,184)
	Cohort average	64% (349,646)	43% (233,265)

Overall, children with any of these disadvantage indicators account for 72% of children not achieving levels 9-4 in KS4 English and maths in this cohort (61% of those not achieving level 5 or above).

The analysis in Table 22 only considers one disadvantage indicator at a time, but we know from (for example) Table 2 that these characteristics often overlap: children who are FSM eligible and/or have identified SEN are also children who are more likely to have contact with children’s services and vice versa.

Table 23 attempts to address this by controlling for CIN 6, FSM 6 and SEN 6 status simultaneously. It shows that these gaps in attainment remain large and statistically significant after we account for correlations between CIN 6/FSM 6 and SEN 6 children though there is some reduction in the size of this gap for CIN and FSM 6 children (Figure 4). Children who have had an open episode with children’s services in the previous 6 years are still 50% less likely to achieve levels 9-4 in English and Maths than those who have not, even after accounting for FSM 6 and SEN 6 status. There is a similar effect for FSM 6 children, while children with identified SEN are around a fifth as likely as those without SEN to achieve this.

Table 23: Fixed effects portion of a regression of CIN, FSM & SEN 6 characteristics on the likelihood of a child achieving levels 9-4 in KS4 English and Maths after accounting for correlations between these disadvantage characteristics

Characteristic	Coefficient (logit scale)	95% confidence interval	Odds ratio
Any CIN in last 6 years (ref = Not)	-0.68	(-0.71, -0.66)	0.50
Any FSM in last 6 years (ref = Not)	-0.62	(-0.65, -0.59)	0.54
Any SEN without a statement/EHC plan in the last 6 years (ref = Not)	-1.67	(-1.7, -1.64)	0.19
Any statement/EHC plan in the last 6 years (ref = Not)	-1.99	(-2.07, -1.92)	0.14

Figure 4: Odds ratios for children with each disadvantage indicator achieving levels 9-4 in KS4 English and maths with and without accounting for inter-correlations between these disadvantage indicators

Figure 5 shows that within the CIN 6, FSM 6 and SEN 6 groups, children who had these characteristics in their KS4 year have lower rates of achieving 9-4 in KS4 English and Maths, compared to children who have had them previously but not in their KS4 year. This suggests a bigger effect of being contemporaneously disadvantaged as opposed to being historically disadvantaged. For example, 26% of CIN 6 children who have an open CIN episode in 2019 achieved this, compared to 40% of CIN 6 children that were not in contact with children’s services in 2019. This pattern is repeated across groups, though the difference is least amongst children who are eligible for free school meals in the last 6 years. Nevertheless, even those with historic disadvantage still have notably lower KS4 results than the cohort average.

Figure 5: Rates of children getting levels 9-4 in KS4 English and Maths amongst CIN/FSM/SEN 6 children split by whether they have their respective characteristic in the most recent time period prior to sitting KS4 exams or not

Children who have been CIN 6, FSM 6 or SEN 6 over a longer time span have notably lower rates of achieving level 9-4 in Key Stage 4 English and Maths. Figure 6 demonstrates that 18% of children who were in contact with children’s services for at least 5 of the 6 years prior to Key Stage 4 achieved these grades compared to 39% of those in contact in only 1-2 of the 6 years. This pattern is similar for both FSM and SEN 6 children.

Figure 6: Rates of children getting levels 9-4 in KS4 English and Maths amongst CIN/FSM/SEN 6 children split by length of time with characteristic identified

Figure 7 demonstrates that children with multiple vulnerabilities have notably lower KS4 outcomes than the relevant group averages. For example, 19% of children who are CIN 6 and had any time on SEN support in the previous 6 years achieved levels 9-4 in English and Maths at Key Stage 4, roughly half the rate of all CIN 6 children (34%). A sizeable group is the more than 50,000 children with both FSM and SEN Support in the last 6 years, of whom only 22% achieved levels 9-4 in English and Maths.

Note: the dot matrix below this chart indicates the broad combinations of characteristics for a specific group. For example a single dot in the FSM6 row indicates all children who are FSM6 regardless of their CIN 6 / SEN 6 status. A dot in both the CIN 6 and FSM 6 rows indicates the group who are both CIN 6 and FSM 6, regardless of their SEN 6 status

Figure 7: Proportion of children with broad combinations of CIN/FSM/SEN 6 characteristics achieving levels 9-4 in KS4 English and Maths. Limited to combinations with at least 500 children in the cohort

Figure 8 repeats the analysis but with groups defined in a slightly different way: a single dot in FSM 6 row indicates children who are FSM 6 but with no other disadvantage indicators. A dot in both the CIN 6 and FSM 6 rows indicates a group who have been CIN 6 and FSM 6 but with no other disadvantage indicators. This graph shows that around 2 in 3 children who are CIN 6 but without any other identified SEN or FSM characteristics in the previous 6 years achieved levels 9-4 at KS4. This is similar to the cohort average of 64%. Among CIN 6 children that have also had SEN but no statement or EHC plan, the pass rate is 28%.

Figure 8 demonstrates just under 1 in 5 children who have an EHC plan in the previous 6 years (but no other CIN/FSM/SEN 6 characteristics) achieved levels 9-4 in English and Maths, compared to 7% of those that have had an EHC plan in the last 6 years and had an open CIN episode during that time (but no other CIN/FSM/SEN 6 characteristics).

Figure 8: Proportion of children with varying combinations of CIN/FSM/SEN 6 characteristics achieving levels 9-4 in KS4 English and Maths. Limited to combinations with at least 500 children in the cohort

Note: the dot matrix below this chart indicates the exact combination of characteristics for a specific group. For example a single dot in FSM 6 row indicates children who are FSM 6 with no other CIN/FSM/SEN characteristics. A dot in both the CIN 6 and FSM 6 rows indicates a group who have been CIN 6 and FSM 6 but with no other characteristics

Conclusions

Overall this analysis echoes previous work suggesting that large numbers of children have these identified vulnerabilities and that (on average) they have notably worse outcomes at Key Stage 4. However, the key addition is that it also demonstrates there is considerable variation within these groups in children’s Key Stage 4 results based on their combinations of these vulnerabilities.

• Children with any of these CIN/FSM/SEN 6 – that is, CIN 6 or FSM 6 or SEN 6 – characteristics represent a substantial proportion of this cohort, accounting for just under 1 in 2 children sitting GCSE exams in 2019.

• They have notably lower rates of achieving levels 9-4 in KS4 English and Maths. Around 4 in 10 of this group achieve at least a level 4 in these subjects compared to a cohort average of just over 6 in 10. Around 1 in 4 achieve at least a level 5, compared to 4 in 10 in the cohort overall.

• Overall, children with any of these disadvantage indicators account for 72% of children not achieving levels 9-4 in KS4 English and maths in this cohort (61% of those not achieving level 5 or above).

• Rates of achieving at least level 4 in these subjects are lowest among those that have had an EHC plan in the last 6 years (12%), and among those that have been looked after at some point in the last 6 years (23%).

• The specific combination of children’s CIN/FSM/SEN 6 status makes substantial differences to their performance at KS4: Nearly two thirds of children that are CIN 6 only – i.e. CIN 6 but not FSM 6 and not SEN 6 – achieve at least a level 4 in English and Maths. This is similar to the average rate for the cohort as a whole (64%) and to children that are FSM only during the 6 years. This compares to 34% of CIN 6 children as a whole.

• Children who have had these disadvantage characteristics identified for longer have worse outcomes at Key Stage 4. However, even those with vulnerabilities identified for smaller proportions of time in the last 6 years have notably lower rates than the cohort average.

• There is substantial variation by local authority in rates of these children achieving at least a level 4 in English and Maths.

  • Rates of FSM 6 children achieving this range from 28% to 68% across local authorities.
  • Rates of CIN 6 children achieving this range from 20% to 54%.
  • Rates for children with SEN support but no EHC plan in the last 6 years range from 17% to 56%.

Taken together these findings suggest that these broad groupings of CIN, SEN and FSM 6 belie considerable variation in terms of KS4 outcomes. This suggests effective interventions to address the broad levels of disadvantage faced by these groups may need careful targeting given these levels of variation. A key question for further work is to examine how interventions may be as effectively targeted as possible to ensure maximum benefit to children with high levels of need as well as to reduce future cost pressures on local areas.

However, even within the combinations examined here, this analysis is limited in the different types of need faced by children it explores (particularly around primary SEN types and/or disability). Useful further avenues for research would be to further explore the drivers of these within group differences, by accounting for a greater variety of needs within these overarching groups such as different SEN types, those with needs identified at a later age, those with experience in alternative provision amongst a few. A final particularly useful angle would be to explore drivers of the variation locally for these children.