Blog: Lies, damned lies, and let's stick it to Hackney

Posted on Friday 4th November 2011

Written by: Chris Game

As you’ll have noticed, it’s National Adoption Week. It’s also Get Local Councils Week, but, given that it’s roughly the 44th of these this year, it may have escaped your attention.  Still, you’ll no doubt recognise the formula. 

Have the PM launch a national, preferably tabloid-tickling, campaign – in this case, ‘Give a Child a Home’. Publish a league table of local authority performance statistics showing, amazingly, that not all 150-plus English councils perform identically. Leak the stats to the BBC’s Today programme, together with a briefing helpfully identifying the ‘poor performers’ and how they’ll be punished. Then, having lit the blue touchpaper, stand back and wait for the outraged headlines to erupt. If you can find a table propped up by, say, Hackney, Sandwell, Brent, Liverpool, or any of the other authorities that Conservative ministers like to pick on for fun, so much the better – which made this week’s story just about perfect.

Of all the different possible measures of a local authority’s record of dealing with children in care, David Cameron has decided that number and, above all, speed of adoptions are uniquely important. There will be a green paper setting out – in the spirit of localism, you understand – new minimum standards for the proportion of children who should be adopted from care each year and imposing time limits on the process. Councils will be named and shamed, and those failing to meet these minimum standards face having their responsibilities taken over by some other authority or agency. And, would you believe it, the council recorded as having over the 2008-10 period the lowest proportion of adoptees placed within 12 months of the adoption decision was, with 43%, Hackney – followed by Brent with 52%. Ministerial Mondays don’t get much better than that.

All such stories based on statistical headlines are difficult to explain and contextualise, especially on the radio at 8 o’clock in the morning, and this was no exception. The headline figures are the ones that stick, and, on the face of it, Hackney’s didn’t look good. They’re for a 3-year period, not a single year; the average for all English authorities is 74%; York achieved 100%, and other 95%+ performers were South Tyneside and Hartlepool, hardly havens of privilege.  However earnestly council spokespersons or Directors of Children’s Services pointed out that adoption is not invariably the best solution for children in care, that speed is only one measure of good adoption practice, and that other measures give very different pictures, you can bet that, if anything was remembered from the story, it was Hackney’s 43%. 

I’m not a statistician, but I was sure that, if I bothered to investigate, what I’d find was precisely what the sector professionals had asserted: there are other measures and they do give different pictures. I very nearly didn’t, but the Government, in its enthusiasm for policy making by league tables, has made these data so exceptionally accessible - http://www.education.gov.uk/a00199753/children-in-care-and-adoption-performance-tables - that I succumbed.  Nothing I found was unduly surprising, but it did confirm my suspicions and expectations, and it seemed sufficiently interesting to be worth sharing.

First thing to note is that the speed of adoption table – the focus of virtually all ministerial and media attention – is just one of 15: 4 on Placements, 2 on Adoption, 4 on Attainment, 4 on Leaving care, and a summary table. Secondly, there is in all tables a lot of missing data – far more than with CIPFA statistics. For the speed of adoption table, for example, which reports 3-year rolling averages, there are no recorded returns at all for 9 of 152 authorities – most, as it happens, currently Conservative controlled.  An additional 32 authorities – among them, Hackney –submitted no 2010 return, but are still included, so the table is actually comparing some 2008-10 averages with others covering 2007-09: not even like with like.

Even if year-to-year fluctuations in the data were minimal, this practice would be questionable, but, as the most cursory glance will confirm, these figures can vary hugely – frequently by 15 points or more – and with them authorities’ league table positions. Taking the 2010 returns alone, the top 4 authorities would be South Tyneside (2nd for 2008-10), Wakefield (4th), Bristol (42nd), and Staffordshire (77th). The bottom 4 would be Bromley (130th) – a touch politically embarrassing – Medway (133rd), Dudley (99th), and Devon (85th).  Equal in 35th place would be Halton (137th), Newham and Kirklees (both 121st), while Sandwell would have shot up from 127th to 32nd, passing on the way Bexley, plummeting from 6th to 47th.  Rolling averages obviously even out extreme year-to-year fluctuations, but to build a policy in such a sensitive field on such variable and incomplete data seems rash, to say the least.

Which brings us back to Hackney, and to Disraeli’s three kinds of lies: lies, damned lies, and statistics. To anyone who would listen, representatives of bodies like the British Association of Social Workers and the British Association for Adoption and Fostering pointed out that it was patently clear, even from some of the other tables published simultaneously, that Hackney’s work with children and families was in many ways something to be praised rather than pilloried.

Take attainment of children in care. Who’s to say if it’s a higher or lower priority than the speed of adoption, but it’s hardly dismissable.  With 47% of the children it looked after continuously for 12 months achieving at least Level 4 in Key Stage 2 in both English and maths, Hackney’s national ranking is 14th.  With 19%, and in 2010 27%, achieving 5 or more GCSE A* to C grades, including English and maths – compared to a national average of 10% - it rises to 4th.

Then there’s the record of children leaving care. The table for the lowest proportion of 19-year olds who had been looked after at age 16 and were not in employment, education or training, is, somewhat surprisingly, headed by Tower Hamlets (also Oxfordshire) with 15%. Hackney, for 2008-10, is right on the national average with 32%, but York, top of the pile for adoption speed, is here down at 137th, alongside Norfolk, with 49% of NEETs. Finally, with 14% of those 19-year olds in higher education – twice the national average – Hackney was up into 10th place, in a table headed interestingly by Barking & Dagenham and West Berkshire with 36%.

I know – it’s getting boring, so I’ll stop. After all, I’ve been doing essentially the same as David Cameron and his ministers: using statistics, as the other famous quote goes, like a drunk uses lamp posts – for support rather than for illumination. The big differences are that I readily acknowledge what I’m doing, and I’m not trying to design a child care and adoption policy.