6 Discussion and Conclusions
This chapter discusses some of the main findings set out in the previous chapters and proposes some implications for policy. It focuses on the changes observed between the two cohorts, and, where different, on the specific patterns observed for children in BC2. In doing so, it addresses the research questions set out in section 1.4.
6.2. Changes in ELC use
The analysis suggested that, on average, 4 year olds in 2014 spent slightly longer in ELC per week than 4 year olds in 2008/09. Although not explored here, this increase might be related to a higher proportion of mothers being in paid employment (cf. Knudsen and Bradshaw, 2017).
The analysis also showed a small but statistically significant increase in the average weekly number of hours children spent in ELC immediately before and after the increase in entitlement to funded ELC which came into force in August 2014. Because of the differences in how data on ELC duration were collected before and after the expansion, this finding should be treated with caution. As noted in section 3.3, the original survey data on ELC attendance had to be converted into a format which was comparable across the different time points (see Appendix A for further details). Nevertheless, the findings do seem to suggest that the increase in the number of funded hours may well have led to an increase in the average number of hours children spent in ELC. It is not surprising that parents who were already using ELC would take advantage of an extra few hours of funded ELC per week. However, given the potential implications for providers, it is perhaps more notable that the increase in hours was visible so shortly after the implementation date (the analysis compared weekly ELC attendance among families interviewed during the six months immediately before the implementation date and families interviewed during the six months immediately after the implementation date).
While the increase in average time spent in ELC was observed for all groups of children considered in the analysis, it was particularly notable among the most advantaged groups - that is, among children living in more affluent households, in less deprived areas, and whose parents had higher levels of education. Indeed, the difference in weekly ELC attendance between the most and least advantaged children increased between 2008/09 and 2014. There are a number of possible reasons for this. For example, more affluent families may have been more likely to have access to providers (in particular, private or voluntary sector providers) which may have been quicker to offer extended and flexible hours that suited their needs, for example because they had already been offering parents extended or wrap-around hours for a fee. It may also be that more affluent parents were more likely to be in work and therefore more likely to take advantage of the offer of additional funded childcare. More affluent and degree-educated parents may also be more likely to believe that attending ELC will be beneficial to their child. The analysis carried out here did not specifically consider the drivers of this increasing difference, though, and the suggestions above are speculative only.
Although children in the most affluent families tended to spend longer hours at their main ELC provider (possibly as a result of attending their main ELC provider for wrap-around childcare, too), in 2014, across each income group children spent an average of at least 15 hours per week in ELC. This suggests that almost all families - including the less advantaged - were taking up their full funded ELC entitlement and many were paying for further hours with the main ELC provider too. Given the current emphasis on ELC provision as a means of closing the attainment gap (Scottish Government, 2016a), this is encouraging for policy makers. Going forward, it is worth considering whether less affluent families would use an even higher number of hours of ELC if they had a better choice of providers and/or flexibility, including the option of combining funded entitlement with wrap-around care. This is particularly pertinent given the planned expansion of ELC and underlines the importance of ensuring equitable access to the increased hours and flexibility for all families, irrespective of where they live or what type of provider they use.
In terms of the type of ELC provider 4 year old children were attending in 2014, the analysis showed a rather similar pattern to that seen for 4 year olds in 2008/09. Namely, that the majority attended a local authority nursery class attached to a primary school. An increasing difference between the most and least advantaged families was evident in relation to parents' choice of ELC provider, with the most affluent families in BC2 more likely than those in BC1 to use a private or voluntary provider. Private and voluntary providers are more likely to offer wrap-around childcare than local authority nursery classes. As such, this finding may suggest that more affluent working parents are particularly likely to prioritise the flexibility of being able to use the same provider for the funded ELC hours and other childcare.
In 2014, at an overall level, 4 year olds were no less likely than children the same age six years previously to attend an ELC provider which achieved top grades across all four quality themes measured by the Care Inspectorate.
The analysis suggested that children living in more deprived areas were less likely than their more advantaged peers to attend an ELC provider with 'very high' or 'excellent' staffing grades. While the proportion of children attending a setting with a high staffing quality grade increased across all deprivation groups between 2008/09 and 2014, it appears to have increased the most in areas with the lowest levels of deprivation. Crucially, though, in 2014 (just like in 2008/09) children living in disadvantaged circumstances were just as likely as their more advantaged peers to attend an all-round high quality ELC provider.
In summary, the analysis showed an increase in the number of hours 4 year old children attended ELC between 2008/09 and 2014, and suggested this increase may well have happened as a result of the increase in entitlement introduced in August 2014. Notably, though, the most affluent families appear to have been particularly likely to take advantage of the increased entitlement. On quality, it seemed that children in the least deprived areas benefitted more than their peers living in more deprived areas from an overall increase in the proportion of children attending an ELC setting with high quality staffing. Nonetheless, looking across the quality themes assessed by the Care Inspectorate, children in disadvantaged circumstances remained as likely as more advantaged children to be attending a high quality ELC provider. Given the association between ELC quality and child outcomes (indicated in this report, and also demonstrated in the wider literature, cf. Melhuish et al., 2015), it is of crucial importance that this equality of access to high quality ELC provision is maintained - and, in the case of staffing quality, improved - as any increase in entitlement is rolled out.
6.3. Changes in children's outcomes upon entry to primary school
Most children were reported by their parents to have adjusted well to primary school - this was the case both in 2009/10 and in 2015. However, children who were in their first year of primary school in 2015 were a little more likely to complain about school and to be reluctant to go to school, and a little less likely to look forward to going to school than those who started school six years earlier. Furthermore, disadvantaged children continued to be reported to have lower overall levels of adjustment than the more advantaged children. That said, although not statistically significant, there were some encouraging signs that the gap in adjustment to primary school between children whose parents had no or lower qualifications and children whose parents had a degree had narrowed. This was driven primarily by improvements among children of the least educated parents.
In 2015 as well as in 2009/10, the majority of 5 year old children were reported as having no social, emotional or behavioural difficulties. Encouragingly, the analysis also showed signs of improvements in the levels of pro-social behaviour and in behaviours related to hyperactivity. However, for pro-social behaviour there were indications that this improvement took place disproportionately among children in the most highly educated households and children who lived in less deprived areas, leading to a slight increase in the gap between children at different ends of these measures.
The analysis showed no change in the problem solving ability of 5 year olds between 2009/10 and 2015. In contrast, there was a slight decrease in average vocabulary ability. This suggests that the improvements in vocabulary between the cohorts found at age 3 (Bradshaw et al., 2015) had, at best, been cancelled out by age 5. More encouragingly, the analysis showed signs of a narrowing of the gap in vocabulary ability between children in the poorest and the wealthiest households. Notably, however, this narrowing appeared to be driven as much by a decrease in ability among children in the most affluent families as by an increase in ability among those in the least affluent.
Thus, the analysis showed some signs of improvements in children's social development on entry to primary school over the period. Conversely, it found no evidence of any improvements either in relation to children's adjustment to primary school or their cognitive ability.
6.4. Is there a relationship between the number of hours children attend ELC and outcomes at age 5? And is there any evidence that the increase in ELC entitlement improved outcomes for children at the start of primary school?
The analysis found no evidence that attending ELC for a relatively small number of additional hours per week was associated with children's outcomes at age 5. However, the findings did suggest that attending ELC for 30 hours or more per week was negatively associated with some aspects of children's behavioural development. Specifically, children who attended ELC for more than 30 hours per week were at higher risk of displaying above average levels of hyperactivity difficulties at age 5 than those who attended ELC for 12.5 hours or less, even when differences in social background were taken into account. Interestingly, the association did not apply for children in more affluent households.  Further analysis is required to understand what drives this apparent difference between children from more and less affluent backgrounds.
It is important to note that this analysis looked specifically at time the child spent at their main ELC provider. As such, the findings cannot be used to draw conclusions about use of formal childcare per se. Nonetheless, time spent at their main ELC provider did make up a substantial proportion of the time children spent in formal care, and the finding is broadly in line with earlier research findings which have suggested that spending very long hours in non-parental care can have negative effects on children's social and behavioural development (e.g. Bradshaw and Wasoff, 2009).
The data for this report were not collected specifically with the purpose of evaluating the increase in funded hours of ELC from 475 to 600 per year, and the results must be treated as indicative in this respect - an evaluation designed to measure such an impact may well have come to other conclusions. That said, as set out above, the analysis did not find any evidence to suggest that attending ELC for up to 16 hours per week rather than up to 12.5 hours per week was either positively or negatively associated with children's outcomes at entry to primary school.
Looking ahead, as set out at the beginning of this report, the Scottish Government have committed to increasing the entitlement to funded ELC to an average of 30 hours per week. The analysis carried out for this report did not find any evidence to suggest that such an increase is likely to have any notable impact on children's outcomes by the time they enter school, either positive or negative. Although the analysis suggested that spending an average of 30 or more hours in ELC per week was associated with a higher risk of exhibiting behaviour problems upon entry to primary school, the differences in risk levels were very small. Notably, though, these findings are based on the current status of ELC provision, where most children experience relatively high levels of quality in the ELC they receive. As demands on providers to offer a higher number of hours increase in line with the increased entitlement there is a risk that quality of provision will fall. Given the association between quality of provision and elements of children's social and behavioural outcomes, a drop in the quality of provision alongside an increase in the number of hours children spend in ELC may well have more detrimental effects than those considered here. This stresses the importance of ensuring that the level of quality of ELC provision does not suffer, even as demands on providers increase.
6.5. Is there a relationship between the quality of the ELC setting and child outcomes at age 5?
The analysis suggested that, compared with attending a lower quality provider, attending a high quality ELC provider may have beneficial effects on some aspects of children's social development - particularly in relation to how they conduct themselves in a social environment and their relationships with others. Specifically, children who attended an ELC provider which achieved at least 'very good' grades across all four quality themes assessed by the Care Inspectorate were a little less likely to exhibit above average levels of peer problems and a little less likely to display below average levels of pro-social behaviour around the time they started school, compared with their peers who attended an ELC provider which did not achieve such high quality grades. Further to this, attending an ELC provider with higher staffing grades was found to be associated with a very small but statistically significant decrease in the likelihood of exhibiting above average levels of peer problems. Notably, for both outcome measures the effect sizes were very small, suggesting that any effects of attending high quality ELC on social and behavioural outcomes are likely to be minor in comparison with other factors such as parenting approaches, peer contact, and health and development (cf. e.g. Bradshaw and Tipping, 2010). That said, the outcomes considered here are very strongly affected by social background characteristics and the quality measures used in the analysis were not specifically developed with a view to informing analysis of effects on children's outcomes. In this context even weak associations are worth taking note of, especially in a context of ELC expansion.
The association found between higher staffing quality and a lower likelihood of experiencing peer relationship problems seems to confirm the role of staffing as an important component of high quality ELC provision - particularly in relation to facilitating positive social and behavioural development in children. Whilst this analysis did not consider different aspects of staffing quality, other research has suggested that staff qualifications and training is likely to play a role, because staff with better (relevant) qualifications can help provide a more stimulating and supporting environment (Scobie and Scott, 2017).
At the same time, as illustrated by the associations with overall levels of quality measured across the four Care Inspectorate themes, our findings also suggest that staffing is not the only aspect of quality that matters. This is also highlighted elsewhere (e.g. Scobie and Scott, 2017) and is important for those seeking to develop high quality ELC environments, as it stresses the importance of viewing ELC quality in a holistic manner.
As noted above, children in disadvantaged circumstances were a little less likely to attend an ELC provider with high staffing grades than their more advantaged peers. Otherwise, however, they were just as likely to attend a high quality ELC provider as more advantaged children. Also, the associations found between attending a high quality ELC provider and social and behavioural development did not appear to be any different for children with different social backgrounds. In other words, more advantaged children did not appear to benefit more from settings which had higher quality grades than less advantaged children. Nevertheless, given the association found between staffing quality and aspects of children's social development, the difference in propensity to attend an ELC setting with a high staffing quality grade between children in more and less advantaged areas is notable, and is something to be borne in mind during the rollout of the planned increase to the funded entitlement.
The analysis did not find any statistically significant associations between either the amount of time children spent in ELC or the quality of their ELC provider and how well they adjusted to primary school.
Earlier analysis using GUS data collected from children born in 2004/05 showed a statistically significant but weak positive relationship between ELC quality and children's vocabulary, indicating that attending an ELC provider with a higher care and support grade was associated with slightly better vocabulary scores at age 5, although the difference in average vocabulary was very small. This report, which uses data collected from children born in 2010/11, found no statistically significant associations between any of the ELC quality measures considered and children's cognitive development at age 5. 
In summary, the analysis found some associations between quality of provision and children's social and behavioural outcomes, although the difference in average outcome was small. In line with much of the existing literature (e.g. Melhuish et al., 2015) these suggest that higher quality environments are associated with better outcomes for children. Whilst the precise findings are slightly different to those found in a previous GUS analysis (Bradshaw et al., 2014), the implications are similar - namely that the quality of ELC provision has a bearing on children's outcomes. This stresses the importance of ensuring quality is at least maintained but ideally improved as part of any proposed changes to ELC provision. Any decrease in the quality of ELC provision as a result of the planned increase in entitlement runs the risk of a negative knock-on effect on children's outcomes.