Examination of Witnesses (Questions 220-239)
CHRISTINE GILBERT
CBE, MICHAEL HART
AND MIRIAM
ROSEN
10 DECEMBER 2008
Q220 Chairman: The 2006 report was
adequate and the 2007 one was quite good.
Christine Gilbert: The JAR overall
was good, was it not? The safeguarding element was adequate. Although
the data used were for the financial year I mentioned, the assessment
would have taken place in the autumn. We are just now finishing
the APAs for this year. The judgment moved from adequate to good.
The assessment and referral of initial and core assessments that
I mentioned were judged to be good and the cases allocated to
social workers were better. A number of positive things were said.
More recently, the Secretary of State asked us to look at Haringey
a few weeks ago. For the broad outline, we used the JAR methodology—our
inspection methodology. However, there was a difference because
it was shorter and focused just on safeguarding. We found some
things, outlined in the report, about issues such as management,
paperwork and practice. However, we also found that some of the
data that we used for the APA the previous year were inaccurate.
For instance, when you looked at the case files, it was clear
that assessments had not been completed, whereas the figures coming
in had reported that assessments had been completed in a particular
time.
Q221 Chairman: So, is this the matter
about which I saw you on television saying that Ofsted had been
misled?
Christine Gilbert: Yes.
Q222 Chairman: Who was responsible
for that misleading? At what level did that happen?
Christine Gilbert: Well, it would
have been the reporting of the data from the local authority.
Another example was the allocation of social care cases to social
workers. In a number of instances, it appeared from the files
that families had been allocated to social workers rather than
the separate children in those families. Therefore, some social
workers would have had a far harder and heavier case load than
was reported. A further key thing was that files were not closed
promptly, as they should have been. That information came in incorrectly
and was fed back into the data profile on performance and so on.
Q223 Chairman: So, Chief Inspector,
with the inspection that rated Haringey as good, did a senior
person from Ofsted speak to the director of children's services
about the overall quality of that inspection?
Christine Gilbert: Of the JAR
or the APA?
Chairman: The one in 2007.
Christine Gilbert: The assessment.
The point is that that was not an inspection.
Chairman: That was the paper-based one.
Christine Gilbert: It was paper-based,
with data and reviews from different organisations—briefings,
as we call them.
Q224 Chairman: So no one from Ofsted
went there and looked at people and talked to people.
Christine Gilbert: There would
have been an on-site meeting, but not an investigation meeting.
That would have been a meeting to discuss what was emerging from
the APA.
Q225 Chairman: Who would have met
with whom for an on-site meeting?
Christine Gilbert: It differs.
Sometimes it would be the director of children's services. More
rarely, the chief executive would be there.
Q226 Chairman: But in this specific
case, you must have a record of who met with whom.
Christine Gilbert: I do not have
that to hand, but it would have been senior people from the local
authority with the two inspectors from Ofsted.
Q227 Chairman: So you do not know
whether the director of children's services was met at that time.
Christine Gilbert: I do not know.
Q228 Chairman: One criticism is that
that particular person was strong on the educational side, but
did not have much background in social services. I will bring
the rest of the team in, but I want to press you on this. What
shocked me, when I looked at the review that the Secretary of
State asked for, was that such a high percentage of the social
work staff were agency staff: it was not far below 50%. When you
come across that sort of figure—even if it is paper-based—do
alarm bells not start ringing? Or is that normal?
Christine Gilbert: I have to say,
that sort of figure is fairly usual across London. Sometimes,
it is not as bad as it looks, because they are agency staff but
they might well have worked there for several years. When we looked
at Haringey recently, a number of them had been in post for at
least six months. They just chose to work for an agency, rather
than a council. But that pattern is fairly typical across London.
Q229 Chairman: Would you not be rather
alarmed if a school was half run by temporary staff?
Christine Gilbert: Yes, this is
an issue. The question of the stability of social workers is really
key; it is the human connection between a social worker and the
child or young person that allows you to see what is really going
on. It is concerning. Absolutely.
Q230 Chairman: But it was not flagged
up in the two reports that we have been talking about.
Christine Gilbert: That is because
it is fairly typical, and it is a well-known issue.
Q231 Chairman: Right, but it does
not come out in the report, so if it was known then, it is a shock
to me, as the Chairman of this Committee. The other thing that
struck me about that was the number of case conferences. One thing
that one knows about human organisations is that, in the IT age,
colleagues who are in the same office, or just across the corridor,
now do not look at each other and speak; they e-mail each other.
I do not know whether that was going on in this case, but the
report that the Secretary of State asked for reveals that many
of the participants in case conferences did not show up. Was that
picked up when Ofsted's inspections—of any kind—took
place? The crucial thing about this world is that it is the facilities
responsible for health and education, as well as the school, the
social worker—the whole team—that know about the child.
This goes back to long before computers—people sit around
and talk about the child and the family that they know. Yet the
evidence that came out, when we looked at this tragedy, is that
some of those people turned up, but some of them did not. Was
that picked up by the report?
Christine Gilbert: That would
not have been picked up in the APA, unless we had been specifically
told that through one of the briefings, which are destroyed. So
it is not clear to us whether that would have been picked up.
I would doubt it very much; in all the APAs that I have read,
I cannot remember that sort of detail ever emerging. It would,
or it might, emerge in a joint area review. The issue that inspectors
found in Haringey recently was that, although there was commitment
to communicate and collaborate across different agencies, they
were almost all working in parallel with the particular children.
We looked at the whole safeguarding arrangements; we only looked
at the case of Baby P in terms of the serious case review—we
were not sent in to look at that.
Q232 Chairman: The most important
point in this whole Baby P case, although there are other cases
like it, is to learn a lesson and to try to do a better job. What
reforms will there be in the inspection process, which will make
sure that we know how many temporary social workers there are
and whether people are turning up at case conferences? That is
surely the level of detail that Ofsted needs.
Christine Gilbert: What has happened
has reinforced the need for inspection—even in between the
three years. It has reinforced the proposals that we sent out
in September. This tragic case has made us go back to those proposals,
and in the end, they will be different because of what has happened.
We were proposing annual visits, during which we will look at
case files, speak to people on the ground and perhaps look at
most of the things that you have just mentioned. I am not sure
whether we would have picked up the sort of detail that you have
just talked about with regard to case conferences without a full
inspection, but we would certainly try to get underneath some
of the data, and we have decided that we will do that. We now
think that a day is probably insufficient time for that.
Q233 Chairman: A day?
Christine Gilbert: We would be
going in for a day, or two days at most. The consultation on those
proposals closed at the end of last week, so we are just going
through the comments that we have got back. For instance, one
of the things that we will be proposing that was not set out in
our proposals in September is that we will send a questionnaire
to social workers in the authority before going in for the day
visits. Those are the sort of things that we are proposing, so
we are really trying to get underneath some of the evidence.
Q234 Chairman: Before I open up the
questioning to other Members, I have just one last question. Some
of the criticisms about your inspections have focused on the general
idea that you can move to a lighter touch, because an authority
is doing all right. From the 2007 inspection report, would not
Haringey have got the lighter touch, because you would have said
that, as it got a good inspection result, you do not have to bother
about it for perhaps three years? Is not that the danger of the
lighter touch? If you do the lighter touch, you only go back to
the cases that you think are causing problems, whereas the real
emerging problem might be in an authority that you thought was
perfectly all right a year or two ago, and you are leaving those
because the lighter touch means that you will say, "Well,
we want to concentrate on the ones we know might be problematic."
Christine Gilbert: That is exactly
why we think that it is really important to use inspection. Even
in the school sector, if we look at what was described as a reduced
tariff, where we essentially went in for a day, we see that those
schools have been very positive about that. We targeted that at
the top 30%, rated according to performance, where all of the
indicators suggested that that will be good or better, but when
we have gone in we have found that over 6% of that 30% are satisfactory
or worse, so we absolutely know that you cannot rely only on data
but have to get underneath some of what it is telling you. The
issue for us is whether the proportionality is in terms of time,
so with schools, we had said initially that we would be going
back to "good" or "outstanding", unless the
data indicated otherwise, every six years. We are making a slightly
different proposal on that now, but that is what we were saying.
We had known, when we reviewed the proposals for what comes after
joint area reviews and APAs, that we had built into the safeguarding
element and the looked-after children element an element of inspection
that would give a snapshot to tell us whether we need to move
the inspection to this year, next year or to year three. We shall
be doing an intensive safeguarding inspection into looked-after
children every three years in every authority in the country,
so that is not proportionate in that way, but we had thought that
the short inspection visits would help us to decide essentially
the priority order of those.
Chairman: Chief Inspector, thank you.
Q235 Mr Carswell: Chief Inspector,
Baby P had not been removed from his home, where he was subject
to continual abuse from his mother, her boyfriend and their lodger,
despite 60 separate meetings with social workers. You rated those
social workers as "good" in 2007 in your performance
assessment. Are not you failing as an inspector?
Christine Gilbert: We think that
the APA has some validity in looking at outcomes, and I think
that most authorities in the country, if not all of them, are
full of people with integrity and commitment, and I do not think
that the data produced by the majority of authorities would have
been faulty. We do not know that, so I wrote to chief executives
on Monday to ask for an assurance that the data submitted for
this current year is fine and that it has been checked and so
on. I do not believe that people up and down the country are submitting
inadequate data.
Q236 Mr Carswell: I have a letter
from Professor Ian Sinclair of York University on the question
of whether it was you or dodgy data. He starts by saying: "First
the tragedy of Baby P did not arise because he was not assessed
or assessed quickly enough." He then goes on to say: "Christine
Gilbert is not complaining that assessments were not done at all
but rather that they were done badly". He goes on: "The
statistics themselves only record whether assessments were in
some sense completed within the time limits and a technical note
warns of `extreme variability in some of the figures' which may
reflect `local differences and interpretation'." Professor
Sinclair continues: "Such ill-defined data can hardly be
regarded as either false or true." Is it therefore not a
little disingenuous of you to pass the buck by blaming dodgy data
when the data does not actually allow you to do so?
Christine Gilbert: I place more
importance on the data than that suggests. A lot is being written
about data being bureaucratic and so on. The data is really important.
It is not just recorded on file. The idea is that someone looks
at it and does something with it, and picks up the connections
across and so on, so that different agencies work together and
focus on the child. It is not correct to say that the assessments
were coherent and sufficiently brought together, because things
got missed in the assessment of this child. We did not go into
Haringey to look in detail at the Baby P case. We looked at safeguarding
arrangements, and at the serious case review. We could see from
that, and from the files on Baby P, that if those files had been
looked at, connections could have been made that were not made,
and that questions could have been asked. Issues about practice
were not picked up. For instance—I will not refer to the
Baby P case here although it is an example of this—the ideal
is that you try to see the child alone or at least have some personal
connection with the child during a session with a social worker.
That is not always possible, in which case you record that. In
some instances, the fact that the child was not—or could
not be—seen was recorded four or five times on file. That
should have been picked up in discussion between that social worker
and the manager in what is described as supervision. Supervision
was happening, but it was not focused on the practice and the
individual child. The data is only there to support the child;
it is the cornerstone of the support.
Q237 Mr Carswell: With respect Chief
Inspector, does that response not suggest that we have now created
an inspection system in which social workers are so busy perfecting
the files, the records and the assessment reporting that they
are not ensuring that vulnerable children are okay. We have created
a Kafkaesque inspection system that does not ensure that the really
important stuff is happening.
Christine Gilbert: Can I be really
clear: inspection has not created these demands. The practice
that I am talking about came out of the Laming inquiry recommendations.
The sort of things that I am talking about were listed there.
It is about what good practice is in safeguarding, and being really
explicit about that. The second point is that you have to manage
that. I agree that it cannot just happen by writing it down on
a piece of paper; that will not help anyone. What you do with
that information and how you use it to support and help the child
is absolutely fundamental.
Q238 Mr Carswell: I have two further
questions. Does this whole episode not show that with your £250
million a year budget, your army of inspections and your gigabytes-worth
of computerised record keeping, Ofsted is doing too much? Your
remit has grown and you need to focus more specifically.
Christine Gilbert: I will be repeating
the points that I made at the beginning. We have done all the
things that each of the separate organisations would previously
have done, but also got added benefit from bringing the four organisations
together. In doing that, as I said to the Committee last time
I sat here, we are reviewing our whole approach to inspection
and regulation to make sure that we and children, young people
and adult learners are gaining the benefits of merger. That is
beginning to emerge. For instance, the work that we have done
on serious case reviews was applying a system and a rigour that
had not been there previously to highlight the issues involved,
making sure that people were aware of them and would be more likely
to do something about them.
Q239 Mr Carswell: On the serious
case review, you know that it remains unpublished. Can we have
confidence in children's services while it remains unpublished?
Do you think it should be published?
Christine Gilbert: This question
was raised with me last week by two leading politicians. I went
back and talked to inspectors about that very question, and talked
to directors of children's services. All of them, to a person,
told me that the reviews would not be done as openly and honestly
if they were going to be published, and that it is really important
for people to be as honest as they can in reviewing what has gone
on. The summary—the overview—is published, and I believe
that our evaluation could well be published. Even if parts of
the letter needed to be redacted, I think you would still get
the general sense. Our overall evaluation could be published,
but I have to say that I am persuaded of the dangers to other
children in the family—other members of the family—involved
with publishing the whole review.
| 
