BBC and sloppy numbers

Without wanting to engage in BBC bashing it must be said that this morning’s BBC Radio 4 Today programme had too many examples of sloppy use of data. There were school boy errors: focusing on the latest figure not the trend, comparing annual data to quarterly data, getting the number wrong, ignoring the impact of inflation when comparing figures over time, choosing the wrong denominator and flooding the debate with large but ultimately meaningless numbers. I am not a regular listener to the BBC’s flagship news programs but I hope they are generally better than this! 

RBS – real terms and the right number

Simon Jack, the business editor, (8.22am, 2.22 in the podcast) discussed the fine given to RBS for its actions in the run-up to the financial crisis. He went on to say that this kicks off the RBS privatisation. Notably he said that the sale of £20bn shares will be the “biggest privatisation in history”.  That’s when my ears pricked up. I think I spot two issues with this claim.

First, he refers to the “Tell Sid” campaign to sell “£1bn” of shares. But, I think the British Gas flotation in 1986 was worth more than that. A House of Commons research paper (page 11 of 14/61) suggests that net proceeds were £7.7bn. That’s a lot more than £1bn.

Second, he ignores inflation. He says that the RBS flotation would be 20 times larger. Well yes, 20 is 20 times larger than 1 but that seems to ignore inflation and rising stock prices. Adjusted into real terms, especially if starting from £7.7bn not £1bn, the RBS flotation is smaller not larger.

There are many ways to adjust for inflation. The FTSE100 has risen over that time from 1600 to 7700, a rise of nearly five fold. Consumer prices have roughly trebled. A BBC article from 2011, marking the 25 year anniversary of the BG flotation said: “Anyone who kept their British Gas shares now owns stakes in Centrica, BG Group and National Grid. These splits make it difficult to calculate exactly how much a current stake relates to the original, but stockbroker Simply Stockbroking has estimated that 100 shares bought for £135 in 1986 would be worth £1,686 today.” This suggests an even larger increase.

This is very rough, but the point to make is that the RBS sale, while large, will not be the same magnitude as the BG sale, and certainly not 20 times larger. Indeed, as the RBS sale will be spread in tranches over a number of years, it does not match BG.

Independent schools and Oxbridge – choosing your denominator 

The programme turned to undergraduate entries to the Universities of Oxford and Cambridge (at 8.42am, 2.42 in the podcast). The piece was hung on a new report from HEPI and interviewed its director Nick Hillman. (Incidentally, Mr Hillman, who was at Christ’s College, Cambridge, was joined by Dr Lucy, graduate of Newnham, Cambridge, and they were interviewed by Nick Robinson ex University College, Oxford. John Humphyrs, the co-presenter, apparently did not go to university.)

The admissions tutor from Cambridge said that their new figures (published today) showed that 64% of first years were coming from state schools. Then Mr Robinson butts in to say that 93% of the population goes to state schools. The implied point is that the 7% of private school kids get 36% of the Cambridge places. That must be unfair!

Alas, Mr Robinson regurgitated the misleading number most regularly used when discussing this old chestnut. In the real world, a good number of state school pupils do not progress to the sixth form. According to DfE table 1a of SFR28/2017, there are about 500,000 pupils in the state school GCSE cohort, dropping to 200,000 in year 12. Another DfE publication (SFR 03/2018) gives more information about the qualifications being followed by sixth formers (including those outside schools) but it’s hard to pull together a coherent picture of achievement.

There are about 95,000 independent school pupils in the two years of the sixth form and about 424,000 in state schools – that means that nearly a quarter of school-based A level candidates are in independent schools. Some will be studying elsewhere. If the numbers are trimmed to those doing three or more subjects, of which at least two are facilitating subjects (needed for the top universities), the proportion of credible candidates that is in independent schools will rise.

So far as I am aware figures that show the number of state v independent pupils with the necessary grades, say A*AA or better, in the right subjects do not exist in the public domain. But we can be sure that Mr Robinson’s 7% for the implied proportion in independent schools is way too low and gives a misleading impression of bias in selection. It won’t be as high as the 36% in the Cambridge admissions but perhaps not far from it. (This is not to detract from the clear disparities between schools when considering the university destination of those students who achieve top grades but perhaps probing the reasons for that is less sexy for the radio.)

We were also treated to the phrase “Ten times less likely”. It’s an odd yet increasing used way to say (I think) that something is one-tenth as likely.

GDP – one quarter of data not the trend

Then we had the MPC meeting (7.14am, 1.14 in the podcast). Kamal Ahmed, economics editor, gave his thoughts. “If you have an economy growing at 0.1%, any marginal change in confidence levels could tip the economy into negative territory …. if the economy was growing at 2%, 2 1/2% ….”. He concluded that the MPC would not raise rates today.

Mr Ahmed is not someone who looks for good news but it’s a shame he makes the data sound worse by mixing quarterly and annual rates. The 2018 Q1 rate was hit by the poor weather which knocked agriculture and construction so it does not reflect the trend. This was widely acknowledged by all reasonable observers.

As the chart from the latest ONS GDP release shows, given the erratic Q1 construction impact, it would not be unreasonable to see the last three quarters as averaging growth of around 1 1/2%. The last two years has seen annual growth of 1.8%, compared to 2.0% for the last eight years since the recovery started (see ONS series IHYR). It might well be that the economy has slowed a bit but it’s not yet dramatic.

Screen Shot 2018-05-10 at 22.55.34.png

Trains – raw numbers mean nothing

The CEO of Network Rail was interviewed about digital railways and was part of a real stats-fest (at 7.56am, 1.56 in the podcast) as he traded figures with Nick Robinson. We heard about 7000 new trains, 4000 passengers waiting for replacement buses at Gatwick, Network Rail being 4th largest land owner, 900 trains hit fallen trees one winter, and the 13 million trees that are next to the 20,000 miles of track. Oh, and there was some confusion about whether it was one million trees or two million that will be felled? And to cap it all, Network rail doesn’t count trees. Not sure who won that but the listener would have been none the wiser.

Advertisements

One thought on “BBC and sloppy numbers

  1. Good points – the private/state school figures are always used, and surely it’s a key responsibility for a expert interviewee in this subject to rebut them if every interviewer is going to use them.
    But I think if you listen back you’ll hear that Simon Jack’s point about the size of the RBS sale being 20x BG was more about the logistics, complexity & thus time scale of the shares being sold, not the value/proceeds.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s