Some notes on the Impact Assessment for the Children, Schools and Families Bill

15 December 2009 by Jennifer

The Impact Assessment[1] and its assumptions seemed to warrant some commentary. Some of my notes are disagreements; others are just questions, where the reasoning behind a statement seems to me unclear.

Please feel free to comment with any corrections or additions. (Thanks to the people who did so on the first draft of this, and contributed to the form it takes here.)

The bit about elective home-based education starts on page 83 of the Impact Assessment. I printed out the relevant pages, highlighted the bits which snagged my attention, and wrote a note on each.

The result is “in order of appearance”, not in order of importance. If you’re short of time, I recommend starting with the following sections:

Point numbers are mine, as the IA itself doesn’t give paragraph numbers.

LAs = Local Authorities. EHE = Elective Home-based Education or Electively Home Educated. IA = Impact Assessment. SAO = School Attendance Order.

— On page 83 —

Information vs ensuring

1. “government intervention is needed to tackle the lack of robust qualitative and quantitative information LAs hold about home educated children, on equity grounds to ensure that every child in England receives a suitable education in a safe environment.”

That’s two goals, with no guarantee that one would lead to the other. “Tackling the lack of information” is one thing, and in some ways not a bad idea. “Ensuring” that “every child in England receives a suitable education” is a different thing. Schools currently fail to deliver a suitable education to lots of children, so why should we believe that LAs or Government know how to do it reliably?

2. “This will:

– enable and require all local authorities to identify all electively home educate [sic] children in their area and ensure they are receiving a suitable education;”

Again: No it wouldn’t, because they don’t know how to “ensure” it.

— On page 84 —

Likely cohort size

under Key Assumptions/Sensitivities/Risks:

3. “We think that the cohort size is likely to be 25,000 to 30,000 and that it is unlikely to exceed 40,000”.

On what basis? This might very well be true, for all we know, but the Badman Report claimed that the “likely” figure was 40k. It seems odd to me that the DCSF would disregard Badman’s figure with no explanation of the discrepancy.

From Badman Report section 6.1:

“Our own data concurred with the DfES (2007) report, that there are around 20,000 children and young people currently registered with local authorities. We know that to be an underestimate and agree it is likely to be double that figure, if not more, possibly up to 80,000 children.”

Mike Fortune-Wood’s research from a few years ago suggested that a multiplier of twice or three times the “official lists” might be about right; applying that to Mr Badman’s 2009 LA totals would give 40k-60k, though I don’t think this estimate is necessarily more accurate than the 25k-30k (see footnote).[2]

The authors of the Impact Assessment also include 80k as a costing possibility, while thoroughly downplaying its likelihood (I think rightly). I’ve not seen any logic suggested for a four-times multiplier, though I note Mr Badman’s rhetorical use of it at the Select Committee:

If I were before you, Chairman, as a Director of Children’s Services and you asked me, “What do you know about the 80,000 children in your care?” and I replied, “I’m awfully sorry, but I can’t tell you very much about them,” I suspect that I would not remain in the post for very long. That, frankly, is the situation in relation to elective home education.[3]

(80k for dramatic purposes and 25k-30k for costings, then?)

If asked for a “likely maximum”, I’d suggest 60k, based on Mike’s research. I can’t see any basis for suggesting it’s more than that. But all we really know is it’s more than about 20k.

— On page 85 —

LAs’ legal duty

4. “LAs will be able to: identify accurately those children that are electively home educated; discharge their duty to ensure that all
children in their area are receiving a suitable education; and satisfy themselves that there are no safeguarding concerns.”

Again we see “ensure that all children in their area are receiving a suitable education”. In this paragraph, it’s described as a “duty”, but as I understand it, this is incorrect.

From the Education & Inspections Act 2006:

13A Duty to promote high standards and the fulfilment of potential

(1) A local education authority shall ensure that their functions relating to the provision of education to which this section applies are (so far as they are capable of being so exercised) exercised by the authority with a view to—

(a) promoting high standards,

(b) in the case of a local education authority in England, ensuring fair access to educational opportunity, and

(c) promoting the fulfilment by every child concerned of his educational potential.

“Ensuring that their functions are exercised with a view to…” is not the same as “Ensuring that all children in their area are receiving a suitable education”. The former phrase describes the way in which they must keep an aim in mind while determining their own actions; the latter describes guaranteeing an outcome.

And if they did have a duty to guarantee that outcome, they’d be in trouble, because (as alluded to above) their track record is of frequent failure, as demonstrated by GCSE results, illiteracy etc. They don’t know how.

Safeguarding concerns

4b. “satisfy themselves that there are no safeguarding concerns.”

This is a naïve conclusion. If detecting safeguarding issues were that simple, Social Services would never get it wrong. The best you can say is that the LA would get one more avenue for identifying possible safeguarding concerns; but that neglects the costs of “false positives”. More on this omission in point 8b.

— Page 86 —

Standard of education as assessed by LAs

5. “LAs tell us that they estimate that 8% of home educated children known to them receive no education and that overall, 20% of home educated children already known to them do not receive a suitable education.”

“LAs tell us”. However, LAs are notoriously unreliable at recognising effective education when it doesn’t closely and obviously resemble school. Therefore, their “estimate” of this number can’t be taken at face value. Some allowance must be made for their current lack of expertise in the matter.

Plus, those stats are wrong anyway, as they included families not yet talked to and those where the family’s only “sin” was to decline a visit to their home, which they are perfectly entitled to do under current law. For details, see Dr Ben Anderson’s review of the statistics.[4]

A correctly added figure for the LAs’ reported concerns would be 7.1% rather than 20%. Some of this 7.1% will then be “false positives” caused by LA staff’s assumptions about “what education should look like”.

The true figure for “no education or unsuitable education” is therefore unknown, but likely to be less than 7.1%. Not 20%.

Likewise, the true figure for “no education” is unknown, but likely to be less than 1.8%. Not 8%.

6. “They [LAs] have no idea about the standard of education experienced by home educated children not known to them.”

I agree. (Quote included for later comparison.)

Child protection implications

7. “there is evidence from serious case reviews and LA evidence that home educated children who are not regularly seen in the community are those where there are most likely to be child protection concerns.”

It’s hard to assess the truth of this when the sources are not cited; but in any case, such a correlation would not necessarily be justification for new laws. The question is, would those children be helped by the proposed system in a way which is not already covered by the current system, supposing that the current system had enough money to fund adequate staff etc?

I’ve not yet seen any case study which showed that (a) the existing system was implemented effectively and yet the child was failed, and (b) the child would have been helped by Mr Badman’s methods. I have on the other hand seen case studies where either the existing system was poorly implemented or Mr Badman’s methods wouldn’t have helped anyway.

Quoting from Dare to Know:[5]

For example, in the case of the foster children of Eunice Spry, the family were visited on a regular basis by Gloucester Local Authority personnel. The police had returned the run-away children to the family even after the children had told them that they were being abused. The problem here was not the law but the fact that the authorities failed to use it.

In the terrible case of Khyra Ishaq, the current law would have been sufficient had the authorities used it to pursue the family more vigorously. How would it help to introduce compulsory monthly visits of all home educators? Resources would have been stretched to breaking point in searching for the needle in the haystack, and a family in such a situation could still simply delay the visit by, for example, simply not being at home.

Uncounted costs

8. “The following assumptions are used in all calculations in this document:

This policy applies to England only;

There are around 20,000 known children in home education, but the total may be 40,000 (likely maximum) or as many as 80,000 (less likely) – costings must be of all three scenarios;

The costs arising from the recommendations will be:

  • the registration scheme – costs to LA

  • deregistration from schools – administrative costs to schools

  • monitoring of the scheme by LA officers

  • additional use of School Attendance orders – administrative and legal costs to LAs;

There are no other new activities, and therefore no other new costs;”

Picking out a couple of items from that:

8a. “deregistration from schools – administrative costs to schools”

It’s not clear to me what these costs would be. The school would have to cross the children off its registers, and notify the LA. Is that all?

Perhaps the authors are thinking of the school’s proposed new duty to assess – but that doesn’t really fit, because (a) it’s not really an “administrative” task but a professional assessment (supposedly), and (b) they say elsewhere that the LAs would reimburse the schools for this, so the wording “costs to schools” isn’t really right.

8b. “There are no other new activities, and therefore no other new costs;”

This conclusion is untenable. I will discuss some costs which have been omitted from the list.

  1. Firstly, there’s the cost of “false positives”.

    By “false positives”, I mean both (a) unfounded concerns of neglect or abuse, which would take the time of Social Services staff (potentially at the expense of children in genuine need), and (b) unfounded concerns of poor education, resulting in a child being forced into school who would in fact have been better off in home education.

    As far as I can see, these are not mentioned anywhere in the IA, let alone having their financial and social impact estimated. This is a very serious omission.

    Does the DCSF believe that annual inspections would generate any “false positives” for either abuse or poor education?

    If yes, what estimates have they made of Social Services staff time in follow-up, and opportunity costs for the child and their family?

    If not, what leads them to that (i.m.o. highly implausible) conclusion?

  2. Although this list does name “additional use of School Attendance orders – administrative and legal costs to LAs”, the authors “decided against” including such costs, asserting that they fall “well within the margin of error”. I’m not convinced. See point 17 below.

  3. Depending on statutory instruments which are as yet unpublished, there may be costs arising from the destruction of child-led education. See point 10 below.

    Even if allowed in principle, child-led educational environments can be damaged by the introduction of third-party inspections, especially when carried out by people insensitive to the dynamics of autonomous learning.[6] The costs in reduced educational success have not been estimated.

  4. It’s hard to put a price on peace of mind, but it’s certain that for some young people, forcing them to meet with LA staff would be a stressful and frightening experience.[7] No costs have been estimated for the reduction in wellbeing (and thereby probably also in educational success) caused by this.

    The effects on some children with autism may be especially devastating.[8]

  5. It’s not clear to me why “opportunity costs to parents for additional meetings” are not included in this list; they’re noted on page 85, so the authors do concede that such costs exist.

  6. Incidentally, opportunity costs to children are not considered anywhere in the document. It would be interesting to know what assumptions underlay this omission.

— On page 87 —

Assessment by schools

9. “school will have to conduct an assessment of the child, which it will then supply to the LA.”

(a) Bear in mind that in many cases, this would be a school which has failed the child. As the school is not an independent party, there is no reason to assume that such an assessment would be reliable.[9] Overestimating the child’s level on leaving school would tend to undervalue the child’s subsequent accomplishments: a plausible cause of “false positives” for poor education, where the child appeared not to have progressed while in fact they had.

(b) The usefulness of this assessment is to some degree predicated on the child’s subsequent education being comparable with the National Curriculum.

Child-led education

10. “We have not yet defined the content or rigour of a “statement of education”, but it is likely to be a short, word-processed document. Exemplar curricula which parents could use successfully are freely available from the DCSF and QCA websites;”

Mr Badman’s report failed to recognise child-led types of education where the curriculum is created dynamically by following the child’s interests throughout the year.[6] In this kind of education, it’s not possible to declare specific goals at the beginning of the year to be measured at the end; the parents’ primary role is to provide a rich environment for the child and support their explorations.

This undefined “content or rigour” is therefore extremely important. Would it or would it not allow families to continue with genuinely child-led education? The potential costs of sabotaging this effective educational environment appear to be entirely unaccounted-for. (I note this above under point 8c.iii, along with other uncounted costs.)

Five assumptions lacking evidence

11. “Every year, 1/6 of the total number of home educated children will leave and enter the database. This assumption is based on the fact that we understand that a high proportion of Home Education decisions are triggered by the child’s move to secondary school—particularly for SEN children, bullied children and Gypsy/Roma/Traveller children.”

A source for the “we understand” would be useful here. I currently have no basis to assess whether this is a reasonable assumption.

12. “Current costings assume that 80% of the 1/6 new cases each year will be deregistrations (and therefore only 20% of the 1/6 will be children who have never been registered with a school). This is based on the assumption that most home educated children have previously been educated in primary school.

We accept that this may not be consistent with the assumption that there are 20-60,000 home educated children completely unknown to the system. However, since the defining characteristic of these children
is that we do not know enough about them, we have chosen to use the existing data. If our information changes as local authorities operate the scheme, we will update our costings.”

If I were in charge of the money for this, I’d be concerned about those two paragraphs! The “assumption that most home educated children have previously been educated in primary school” doesn’t appear to be based on anything whatsoever, and as far as I can see, the second para is the authors admitting that. (In other words, “We don’t know how much it’s going to cost, but we’ll find out when we actually do it.”)

13. “We think it is reasonable to assume that a child with SEN will often warrant more in-year monitoring.”

What’s the basis for saying that?

— On page 88 —

14. “The newly identified children are the ones who are most likely to require additional monitoring”

Based on what evidence? In my experience, people who remain off the LAs’ lists succeed in doing so at least in part because no-one who knows them has the slightest concern about their children. Otherwise people get “shopped” to SS, and end up on the official lists. So this assumption doesn’t make sense to me.

15. “Local Authority Training

The estimate is based on a slightly higher number of staff than we believe will actually be involved with registration and monitoring, and is based on training needing to be redelivered every 3 years.

Every local authority is different, using different kinds of staff to deal with Home Education (usually either EWOs or officers specifically recruited for the Home Education role), and each has a different rate of
staff turnover. So, a 3-year rolling programme of training is an appropriate assumption to ensure that training funding is adequate.”

“Is an appropriate assumption”: Is it? Based on what evidence?

Mistaken equation of EHE with “school at home”

16. “Opportunity costs to parents

The opportunity costs to parents of meeting with local authority officers have been factored into the costing. However, we have not included a cost for the preparation of an education plan on the basis that:

  • Even though parents and carers may not give it that name, it is a core part of planning ahead to deliver home education for their children. Any change will not represent additional time invested, but instead mean that parents and carers are using some of the time they devote to home education differently.

  • Curricula are available for immediate download from QCA and DCSF websites, and are adequate for the purposes of education planning.”

Homing in on a couple of those bits:

16a. “Any change will not represent additional time invested, but instead mean that parents and carers are using some of the time they devote to home education differently.”

In your dreams! This is so wrong, I could write a whole essay on it (and may do). But it’s completely congruent with the authors’ evident belief that home ed is basically “school at home”, and is planned within the same frameworks as school teaching would be.

16b. “Curricula are available for immediate download from QCA and DCSF websites, and are adequate for the purposes of education planning.”

if you’re doing “school at home”.

Cost of School Attendance Orders and appeals

17. “We considered including the cost of SAOs in our calculations, but have decided against it for the following reasons:

  • We have no direct data on the number of SAOs used by local authorities each year.

  • We can, however, use a proxy measure—Ministry of Justice figures on the number of adults sentenced for child truanting offences.

  • The actual number of children of 5-16 in school in 2006-07 was 7,440,000. So, the prevalence of sentencing was approximately 0.05%.

If we apply this percentage to 40,000 children, this means that around 20 children would be affected and for the (highly unlikely) prospective cohort of 80,000, this is 40 children. Divided among 150 local
authorities, this is well within the margin of error.”

Apologies for the length of this bit, but it just isn’t that simple…

  1. First of all, it’s important to recognise that SAOs under the new regime would have a different legal framework from the existing kind, and could (or must?) be issued for shortcomings in paperwork as well as genuine educational reasons.

    (3A) If it appears to a local authority in England—

    (a) that a child of compulsory school age in their area is a home-educated child, but is not registered on their home education register, and

    (b) that it is expedient that the child should attend school, the authority shall serve a school attendance order on the child’s parent.

    (3B) In determining for the purposes of subsection (3A)(b) whether it is expedient that a child should attend school, an authority shall disregard any education being provided to the child as a home-educated child.[10]

    In other words: When the LA finds a child not on the register, the LA may not take into account whether the child’s current education is serving them well, only whether it’s “expedient” for the child to attend school. “Expedient” has not been defined, except to rule out consideration of the child’s current education.[11]

    Betsy Anderson has done some analysis of Schedule 1,[12]
    and drawn up a list of things which, it appears, the LA could use as reasons to issue SAOs:

    Under the proposed Schedule 1:

    • Whilst not legally obliged to, if a parent does not apply immediately for registration on the authority’s home education register the child can automatically receive a school attendance order.

    • If a parent is educating brilliantly and the child has 20 GCSEs, but is not on the register, the child can receive a school attendance order.

    • If a parent does not submit a plan for the next year’s education on time, or in sufficient detail, or in the authority’s designated format, the child can receive a school attendance order.

    • If parents do not adhere to their submitted plan, the child can receive a school attendance order. It is irrelevant whether the education provided was an improvement on the original plan.

    • If a family’s circumstances change during the year, the child may receive a school attendance order.

    • If a parent applies to register to stop the school attendance order process, the process stops. But then the authority can refuse to enter the child onto the home education register, and serve another school attendance order.

    • If an authority once denies an application to register a child, it will have continuing power to deny any future applications concerning that child.

    • If the parents object to the local authority entering their home on a routine visit, the child may receive a school attendance order.

    • If parents object to the authority questioning their child on his or her own with no parent or carer present, as Schedule 1 allows, the authority can revoke the child’s registration and serve a school attendance order precisely because the objection was made.

    • If the parents do not comply with any school attendance order entered according to Schedule 1, they may be guilty of a criminal offence.

    • The Secretary of State will be empowered to impose further regulations and technical requirements, beyond parliamentary debate, setting even more conditions upon parents’ ability to home educate.

    Now, if the child’s best interests are put first, and if their home-based education were serving them perfectly, then obviously one would not actually want to force them into school. Yet, what’s the point of issuing a SAO if you don’t actually want the child to end up in school?

    Seems to me it makes sense only in the context of enforcing compliance. In other words, School Attendance Orders would no longer be about believing (and potentially explaining in court) that school was the best place for the child; they would be functioning as a threat to the family. “Cooperate with our interference in every way, or else.”

    Let’s face it, criminalising the parents directly for wrong paperwork was never going to “play well”. As a substitute method of enforcement, this is brilliantly ingenious – because a significant number of home ed parents would rather go to prison than send their child back to a school which had harmed them. (Though if you don’t comply with an SAO, then technically your children are truants, even if in fact their education is proceeding marvellously, and you can still end up in prison.) So yeah, very clever.

    But as an expression of genuine commitment to the children’s wellbeing, it makes no sense at all. Sending children back to school just because the paperwork was wrong, even though they were happy, learning and thriving? Does anyone think that’s a good idea? I can’t really imagine that most LA staff would think so (even if some might like to have the threat “in reserve”).

    So it’s very hard to predict at the moment how LAs would in practice use these new “paperwork-justified” SAOs. Would they scatter them around like confetti at the slightest hint of non-cooperation or untogetherness, then withdraw them equally speedily when the parents fall into line? Or would they save them up for the really stubborn parents who have fundamental conscientious objections to the new regime? I suspect that in fact nobody really knows yet.

    (I also suspect it’d be a “postcode lottery” &/or “individual-staff-member lottery”, as LA-to-EHE-family relations already notoriously are.)

    But what’s for certain is that SAO rates in the new regime cannot be estimated by reference to the current rates. The new SAOs would be a different beastie from the old ones, and would be used in different ways.

    And yet that does seem to be precisely the assumption which the authors of the Impact Assessment have made.

  2. Actually, even if the basic premise were sound, they have the wrong figure for the current rate. They say “We have no direct data on the number of SAOs used by local authorities each year”. But if they mean SAOs issued in respect of children in home-based education, then actually we do have some data now available, thanks to FOI requests from the home ed community.[13]

    141 LAs answered the FOI requests, and altogether, they had issued 77 SAOs in the year 2008-2009.

    This means that for all EHE children in England that year, the figure was probably something near 82 SAOs.[14] (Not 10, as the authors’ “proxy measure” would have predicted.)[15]

  3. But there’s another reason, also numerically significant, for why current SAO rates can’t be used to predict what would happen after new legislation.

    You see, under the current regime, LAs don’t necessarily issue SAOs even when they claim to think the education is inadequate.[16] The reasons for this are unclear as far as I know, but seem to include LA staff’s unfamiliarity with current law, and the expenses involved. And this all comes under the general heading of “LAs complaining that they don’t have enough powers” – pretty much the entire premise of this legislation (albeit not so far borne out by any evidence).

    LAs currently claim to believe that about 360 EHE children in England are receiving no education.[17]

    (This is children already on the LAs’ official lists; no guesswork included about the educational status of unknown families.)

    So if the LAs got the new powers they were asking for, shouldn’t that mean that they finally get on with sorting out these families? Shouldn’t that mean that in the very near future after any such new legislation, they’d be issuing 360 SAOs?

    (I’m not saying those 360 children necessarily aren’t receiving any education; as I said under point 5, there’s the issue of LAs’ lack of expertise in assessing non-schooly environments. But we’re talking here specifically about what LA staff believe, since that’s what would drive the number of SAOs.)

    And what about the other 1,060 or so children whose education is allegedly “unsuitable”?[18] Let’s imagine that for half those children, their parents do manage to satisfy the inspectors; that would leave about 530 children still heading for SAOs… wouldn’t it?

  4. Getting back to the paperwork-compliance SAOs: as I’ve said, we don’t really know at all how LAs would deploy these in practice.

    Would it be reasonable to guess 5 paperwork-related SAOs per LA in the first year? Would it be reasonable to guess 50 for some LAs in the first year, given how many children would be coming to their notice for the first time?

    Let’s take a very conservative guess of 3 per LA; that would be 450 for all of England in the first year.

  5. Adding the three estimated figures described above, we can construct some kind of vague estimate of how many SAOs might be issued in the first year of the proposed new regime.

    360 “no education” + 530 “unsuitable education” + 450 “paperwork-compliance” = 1,340 SAOs in the first year.

    Remember, the first two of those figures are based only on children the LAs already know about, so this is a pretty conservative estimate all round.

    For subsequent years, I’d guess fewer education-related ones (on the basis that some families unable to satisfy the inspectors would have been forced out of EHE), but not necessarily any fewer paperwork-related ones.

    I’ve “shown my workings”, so it’s pretty clear that we don’t know what on earth would really happen. But I don’t think that guess is any less plausible than the extremely implausible forecast of 20 to 40.

    I don’t know what it costs now to issue SAOs; and it occurs to me that the cost of issuing a new-style EHE-specific SAO could be different anyway from the cost of issuing the existing type, for instance because there was no need to provide the same kinds of evidence in court… if that’s what the new legal framework would imply. So I can’t say what this means in money terms. But I’d like to see at least some attempt to estimate it separately (as opposed to submerging it under “margin of error”).

  6. A serious omission in this Impact Assessment is the cost of forcing children into school whose home-based education suited them perfectly, but whose parents had failed to complete the correct paperwork. It’s inconceivable that forcing children into school for this type of reason would always be the best decision for the child’s wellbeing and education. The IA currently includes no estimate of the costs in such cases to the child’s wellbeing, educational success and future notional earnings.

    [I.m.o. the use of SAOs in this way would be ethically indefensible, due to its intentional disregard of the best interests of the child. But I limit my note here to its neglected implications for the Impact Assessment.]

  7. Then there’s the costs of appeals.

    It seems likely to me that most of these SAOs would be against the parents’ wishes.

    It’s unclear at the time of writing how parents might challenge a new-style EHE-specific School Attendance Order.

    • Betsy Anderson’s analysis of Schedule 1 suggests that the courts would have no role in this.[19] It’s not clear what the appeals process would be if not via the existing legal system.

    • It’s not clear whether the body appealed to would or would not be allowed to consider the child’s education. (As we’ve seen, the LAs are specifically instructed not to.)

    If SAOs could still be challenged on the grounds that the child was better served by their existing home-based education, then I would expect many to be so challenged. Appeal costs would include LA staff time in preparing their evidence, parents’ lost time, and the costs of running the appeals body.

    If SAOs could not now be challenged on the grounds that the child was better served by their existing home-based education, then the number of unjust and damaging forcible returns to school would be all the higher.

A completely unsupported assertion

18. “LAs tell us that home educators who avoid interaction with the local authority tend to be providing inferior education.”

This is a ridiculous thing to say. How would they know? I refer the reader to quote number 6 above.

— On page 89 —

Imagined theoretical benefits

19. “According to DCSF calculations based on the Labour Force Survey, moving from not obtaining any qualification to achieving 5+ A*-C GCSEs would increase a child’s lifetime productivity by £186,500.”

It doesn’t follow. GCSE results in that context are largely a marker for the roots of eventual earning power, such as parental support and natural aptitude for academic subjects. It’s correlation, not causation.

Only in rare cases are the actual GCSEs essential to a particular critical path to success; we know that home ed children frequently succeed in moving to employment or higher education using a “life portfolio” approach, and indeed are often sought-after candidates.

The real question is what quality of education the children are receiving, and to what degree the LA’s intervention is capable of improving that versus undermining it. Point 1 is again relevant here.

(This whole section also implies an ethically questionable assumption about earning power as the only relevant measure of success; perhaps that’s inevitable in the context of such a money-oriented document, but let’s at least acknowledge that not everyone aspires to the jobs which pay highest.)

20. “If 1600 home educated children do not achieve good GCSE results”

Note the “If” and see point 5.

21. “Another 2,400 home educated pupils are receiving inadequate education”

… According to LAs. This is not a fact; it’s the opinion of ill-informed people, exaggerated by faulty stats. See point 5 above.

Summing up on quotes 19 to 21: It’s far from clear that anything like that number of home ed children are IN FACT receiving an unsuitable education (as opposed to “in the opinion of ill-informed people, exaggerated by faulty stats”); and it’s far from clear that the LAs’ interventions would have a net positive result.

In short, ALL the millions of pounds of theoretical financially-quantifiable “benefits” imagined in this part of the document are underpinned by nothing much.

Again, see the work of Dr Ben Anderson, cited above; this indicates that even simply correcting the wrong sums would be sufficient to alter the DCSF’s prediction from millions of pounds gain to a likely net loss over 10 years.

When rates of ‘no’ and ‘unsuitable’ education that reflect evidence actually presented to (or certainly available to) the DCSF from the review it commissioned and from a subsequent supplementary data collection are applied to the DCSF’s own modelling assumptions, there is either a small net ten year quantifiable gain or (more likely) an overall net loss.[4]

(emphasis in original)

— On page 90 —

Two fictional statistics

22. “We do know, however, that post compulsory education, home educated young people are 4 times more likely not to be in education, employment or training than other young people.”


In fact, we “do know” that this statistic is fictional, due to glaring errors in data collection.[20]

Issues include:

  • More than one definition of NEET (Not in Education, Employment or Training), was used within the figures, and the various definitions are not comparable.

  • There is no reason to assume that the “officially listed” children are representative of those not known to their LAs and hence not counted.

  • After “school leaving age”, LAs have no duty to monitor outcomes and parents have no duty to keep LAs informed; therefore, even if an LA did know that certain EHE children existed, they still don’t necessarily know what they did afterwards.

  • The LAs’ figures included estimates (a.k.a. guesswork), and Mr Badman did not note how much was guesswork and how much was actual counting.

  • The LAs who replied to Mr Badman on this question were self-selected.

In fact, there is no reason to think that the NEET figures for EHE children are any different from children as a whole.

23. “Recent data collected from local authorities indicates that the percentage of EHE children subject to a Child Protection Plan is 0.4%. The total number of ‘other’ children who were subject to a CPP plan is
0.2% We can therefore say EHE children are 2 times more likely to be subject to a CPP than other children.”


The inclusion of this statistic ought to be cause for serious concern, since its fictional status had been demonstrated at the Select Committee Enquiry over a month before this Impact Assessment was published.[21]


[1] Copies of the Impact Assessment are available from Look for a “download” button in a box entitled Children, Schools and Families Bill – Impact Assessment PDF.

[2] Mike F-W’s research: As I understand it, Mike asked home educators via email discussion groups whether they were on or off their LA’s official lists. At that time, around 2002 to 2005, about half to one-third were on the official lists.

The main problem with this method is that EHE families on the internet may not be on their LAs’ lists at the same ratio as families who aren’t.

If I had to guess one way or another, I’d guess that families off the net were more likely to be on the official lists, as they’re perhaps more likely to be under the illusion that registration is already compulsory. Therefore, if the true figure falls outside the range suggested by Mike’s multiplier figures, my guess would be a lower total figure for EHE children, rather than even higher.

[3] Mr Badman’s rhetorical use of the phrase “80,000 children”:, the transcript from the session on 12 October 2009. See question 2.

[4] Dr Ben Anderson’s review of the statistics:

[6] Autonomous learning: A good basic explanation is in Autonomous Education UK’s submission to the Select Committee Enquiry. If you’re reading online, I recommend the copy with working internal links. The Government’s officially published copy (no hypertext) is at

[7] “Stressful and frightening”: A survey of children conducted by Ann Newstead in June 2009 asked children whether they thought it was right for LA staff to visit them in their homes. (Note that these were not responses to the idea of being seen alone; that was a separate question.) Responses included the following:

“I am very scared about it….I don’t want to talk to them even with my Mum there, it’s scary. it makes me sad. I am 6. We talked to them before, my mum sent letters, they didn’t have to come before. I liked it better that way. My cousins at school don’t have them come to their house. So it’s not right, and I don’t want it.”

“I do not want to speak to anyone who I do not know and who might make me go back to school.”

“The lady was horrid where we live before. Mum cried after she had gone.”

“I would also worry that they might try to force me back to school.”

Summary of survey results:

[8] Children with autism: Gill Kilner discusses this, and quotes some first-hand accounts from parents, at The Equality Impact Assessment under discussion in that article is part of the same set of documents as the (“ordinary”) Impact Assessment which I’m analysing here.

[9] “School’s estimate perhaps not reliable”: I don’t mean that people would deliberately lie (except in rare cases); but where there’s a role for subjective opinions, it would be wise to allow for bias. Human beings want to believe they’ve done some good. “Saving face” and similar social/psychological motives can also play a part, especially where the child’s progress at school has been strikingly poor or backwards. Conversely, I can’t think of any incentive for school staff to play down the child’s accomplishments while there. So I’d expect cumulative errors to tend towards the wishful-thinking direction.

[11] “Expedient”, undefined: Possibly this word has been included to allow for the situation where the LA doesn’t actually have a school place to offer the child. In other words, it would be “expedient” for the LA not to have the problem of finding a school place.

[12] Betsy Anderson’s analysis of Schedule 1, including a list of potential reasons for SAOs: This extract is taken from the last section of the Summary.

The page cited at footnote [10] may also be useful.

Alternatively, see the Government’s “explanatory notes” for the Bill: However, these are rather impenetrable, due to the amount of cross-referencing with other bits of law not quoted on the page.

[13] SAOs, available info from FOI requests, summary by Tania Berlow et al: Note that Firefox and Opera browsers may have trouble with Google Docs; IE or Google Chrome are more likely to work.

[14] 82 SAOs: All I’ve done here is allowed for another 9 LAs who didn’t reply to the question, assuming their SAO rates were similar. It’s not an exact figure, which is why I said “something near”.

[15] “Proxy measure”: It’s not clear to me why the authors ever supposed that truanting offences could be used as a proxy measure for SAOs. They are quite different phenomena.

[16] “Don’t necessarily issue SAOs”: This is immediately evident from a comparison of the SAO figures with the claimed “no education” figures. SAOs for the year, around 82; “no education” currently around 360 (both figures explained in nearby paragraphs in this article).

[17] 360 children: 1.8% of 20,000. That’s using the correctly added figures from point 5, rather than the wrong ones originally reproduced in the Impact Assessment.

[18] 1,060 children: 5.3% of 20,000. We have 7.1% for “no education or unsuitable education”, and 1.8% for “no education” alone; the claimed “unsuitable education” percentage makes up the difference, 5.3%. Again, see point 5, and Dr Ben Anderson’s review as cited in footnote [4].

[19] “Courts would have no role”: see again the pages cited at footnote [12]. The relevant sections in Betsy Anderson’s analysis are 3.4.6 and 3.10.

[20] Errors in NEET figures: The list of points which follows is based on the work of Tania Berlow et al. See (already cited under footnote [13]).

[21] Fictional status of CPP statistic, discussion at Select Committee Enquiry:, the transcript from the session on 12 October 2009. See questions 13 to 18. There are other faults in this statistic; Graham Stuart here identifies only a particularly glaring one.

Here, have a linky index…
Top of document
Information vs ensuring
Likely cohort size
LAs’ legal duty
Safeguarding concerns
Standard of education as assessed by LAs
Child protection implications
Uncounted costs
Assessment by schools
Child-led education
Five assumptions lacking evidence
Mistaken equation of EHE with “school at home”
Cost of School Attendance Orders and appeals
A completely unsupported assertion
Imagined theoretical benefits
Two fictional statistics

Appreciation, criticism & new ideas all welcome...

Optional HTML tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Comments are moderated. Please no name-calling; please speak for yourself from your own life, or else say where you got the info. Thanks :-) Why these commenting guidelines?