Sat Jul 7th, 2012 at 04:09:56 AM EST
We have now concluded the first three modules of the Leveson Inquiry into the standards of the press. Throughout the recent days there have been many complaints that Leveson has become a partisan witch hunt, when questioning has concentrated on the Murdoch BskyB deal that was in process at the start of the current session, it has necessarily dealt with those in power at the time, there has also been an allegation that there has been nothing coming out of Leveson, as nobody can remember anything. How does that relate to the figures?
When we look at the evidence, transcripts of witness sessions run to in excess of 450,000 lines of text with approaching 30,000 questions asked by lawyers at court 73 and the judge. Covering in excess of 500 hours of time, spread over several months. Now in previous times this would have just headed off to sit on an academic shelf somewhere, but with the internet, and computing power, nowadays, anyone can download the lot, and as the saying goes, there's always someone out there with an idea and time on their hands.
My idea turned up in a conversation with Brit from DKos, during one witness's evidence, about the number of times there had been mention of "I don't remember". Which lead to the idea of throwing all of the evidence into several spreadsheets, and seeing if there was anything interesting that came out at the far end.
My figures actually come out lower than several others that have been quoted on the internet, other assessments have just counted the number of times phrases have occurred whereas, firstly I've looked at the context, so "I just don't remember" gets counted, whereas "can we come back to that later, in case I don't remember" wouldn't. Secondly if a person says "I don't remember" more than once to emphasise a single denial of memory in answer to one question, then that only gets counted as one occurrence. Module one and two were dealt with rather more sparsely than module three, to make the task even remotely practical, searches were conducted using software, but individual segments of the results were checked by hand to confirm that results were consistent. Module 3 however, and major figures were separated out and searched through entirely by hand.
Module one consists of the evidence given by general members of the public, celebrities and lower level reporters and consists of approximately 18,800 questions, the people questioned are spread over the age range and if anything biased towards earlier in the period of events so you would expect a higher rate of forgetfulness. The figure actually comes out as 58 episodes of forgetfulness, across the whole set of witnesses. (Remember that number, it'll come in handy later, and you just know with this warning that coincidentally it'll turn up again).
So with an average group of citizens, picked by a selection of newspapers and a committee, we have a figure for average number of questions that they don't remember. And that figure is 0.3%, so for every 300 questions that the person on the street is asked about themselves, or things they have been involved in, there is one they can't remember.
Module 2 and a bit
The second module and the third's witnesses sort of bleed into each other or rather they were the last group I ran through, and by this point staring at spreadsheet cells, my eyes felt like I'd been rubbing sand into them for months. So I didn't spend too much time searching for the group boundaries. Basically this group is those people who were either in the police section, or were newspaper proprietors. If you were a major figure in politics or the ongoing scandal then you were plucked out for the third group. (If you feel you were improperly consigned to a more minor group, feel free to get in touch and I will study your evidence more closely if you can convince me of your lack of memory, or wish to pay for my time :) )
Module 2 and a bit consists of just short of 11,000 questions, however the witnesses, mainly being people you'd expect to take notes and be able to remember events for court cases and to publish in their newspapers, managed to only forget on 23 occasions. Giving them a figure of roughly .21% forgetfulness or roughly one failure in 475.
Module 3 and a tiny bit
People who got into this group got searched through entirely by hand. To get into it, you had to be either a major UK politician, a major Murdoch member of staff, or someone closely involved in the Jeremy Hunt BskyB line of questioning and in total it's a 3 Mb spreadsheet just for them, so you can see why other people were searched somewhat more mechanically.
The thirty witnesses in this section were asked a grand total of 4,212 questions, and between them failed to remember 3.3% of the time. Even though some of the subject of their questioning was a matter of months away, rather than somewhere roughly five years ago for many of the module one witnesses.
There are several factors that seem to increase your loss of memory:
Being a politician 2.48% average 8x worse than the average person
Being Prime Minister 3.77% average 12x worse than the average person
Being involved in the Jeremy Hunt affair 5.94% nearly 20x worse than the average man on the street
Which we can split further
Being a senior Murdoch Employee 5.77% 19x worse
Being on the Government side in that deal 6.74%
So government ministers, and their civil servants and special advisors involved in the Hunt affair have memories that are over 22x worse than the average person. One government witness on their own had 58 memory failures, as many failures of memory as the total in the whole of the first module. or a rate of one memory failure every fifteen questions.
Now if we split it down into individuals and show it as a graph:
Note the red line at 4%. If you are a senior member of the Murdoch organisation, or were involved in the BskyB bid from the side of the government, you fall above the 4% line, if you weren't involved in any way with the Hunt affair or didn't work for Murdoch in a senior capacity then you fall below that figure. (Interestingly Rebekah Brooks appears to have the best memory of all the senior Murdoch staff called, a figure that can't help but keep people awake at night as her trial date sneaks closer.)
Of those who did particularly badly with remembering "the whole truth" we have, in order of forgetfulness:
Rupert Murdoch (4.59%)
Jeremy Hunt (5%)
Colin Myler (5.1%)
Tom Crone (5.3%)
David Cameron (7.1%)
James Murdoch (7.2%)
Andy Coulson (9%)
Adam Smith 58 Failures (11.4% memory failure rate)
How these figures fit with yesterday's claims by Jeremy Hunt that he feels completely vindicated by his appearance before the Leveson Inquiry is something that needs looking at closely, and his claims treating with some scepticism.
So are the results significant? Does a figure for memory loss that is nearly 25 times that of the average person inspire confidence in a Prime Minister? Does such a marked difference between those involved and those not involved suggest that they were entirely honest in their replies? And can you say that in accordance with their oaths they were answering, "The truth, the whole truth, and nothing but the truth?" It seems to me, that anyone involved appears to have a major problem with "the whole truth" as a concept. But how this view of the evidence given ties up with that viewed by Lord Justice Leveson will only be seen when he produces his report at the end of the Inquiry.