As a result of the bee in my bonnet, you’ll now be painfully aware that clause 11(2)(a) of the Railways (Accident Investigation and Reporting) Regulations 2005 obliges RAIB to complete and publish its reports “in the shortest time possible and normally not later than 12 months after the date of the occurrence”. You will also have gathered that, in my view, the Branch misses this deadline far too often, compromising the wider industry’s ability to learn lessons quickly and effectively from serious events.

Over the past 12 months, the Branch has published 27 reports into incidents on the heavy rail network. On average, these have taken 390 days (12.8 months) to complete, with 12 surfacing after the regulatory deadline. Another five dived under the bar within a week of it.

In November 2008, I conveyed my concerns to Anders Lundstrom, head of the European Railway Agency’s safety unit. His is an organisation with no enforcement teeth so he pointed me in the direction of RAIB and the Department for Transport. He did though concede that the Agency had noticed “substantial delays” in respect of the deadline and “there should be justified reasons if the deadline is not met”.

So my next port of call was Carolyn Griffiths, the Branch’s Chief Inspector, of whom I asked two simple questions: what causes these publication delays and how are they being resolved? She claimed that RAIB’s inquiries are broader and go into greater detail than those formerly chaired by RSSB, pointing towards Grayrigg as an example. This event could hardly be described as ‘normal’ so it was of no great surprise that the inquiry went on for 20 months. Her choice of this example to generally explain the Branch’s timescales smacks of fog-wafting and I remain wholly unconvinced by her argument. And it doesn’t explain why the unexceptional occurrence near Victoria Station in November 2007, involving a trackworker being struck by a train, took even longer to investigate than Grayrigg did!

Ms Griffiths declares that the Branch is “always looking to further develop and improve our operations” but no specific actions are detailed in support of this assertion. In her response to a follow-up email, she maintains that “there are investigations that are completed within the year but I think it is unrealistic to think that all will be. The legislation states an expectation for delivery”.

That’s a hugely arrogant position to adopt in relation to the law. Since when has it conveyed ‘expectations’ rather than obligations? Is the speed limit just an expectation too? Who redefined the English language? The regulation is clear, stating that “The Branch shall report…not later than 12 months after the…occurrence”, not “The Branch can report…” or “The Branch might report, if it can be bothered…”

And so onto Lord Adonis who, back in March, was the Minister of State for Transport. His reply to my first letter was acutely fur-coated, side-stepping my questions with the fleet-footedness of a gazelle. It was clear that RAIB had spoon-fed him some sickly syrup which he duly regurgitated. Undeterred I tried again, prompting a second response in which he claimed that the 12-month requirement needed to be “put into proper perspective”. Perhaps 12 months becomes two years if looked at from certain angles? According to his figures, the Branch’s mean overall inquiry timescale was 10.6 months, with a modal average of 12 months.

We can all manipulate statistics to support our case and RAIB does this elegantly through the inclusion of inquiries into mishaps on heritage lines and tramways which, for some reason, it seems able to conclude in two minutes. This distorts the overall figure to its advantage. The Branch simply won’t acknowledge the validity of looking separately at our main line rail network on which 90,000 people are certificated to work and passengers hurtle through the countryside at 125mph.

Although Adonis had slammed the door shut, he had let the cat out regarding the Ipsos Mori survey. Unsatisfied with the two-page digest lurking at the back of RAIB’s 2007 Annual Report, I made a Freedom of Information request for chapter and verse. This was turned down and, in some ways, I’m very thankful it was. I would not otherwise have been able to enjoy the ludicrous barrel-bottom scraping that churned up grounds for non-disclosure. Ms Griffiths was clearly very keen that I should not immerse myself in the report’s full depth.

“The statistics represent the views of persons in 2007 and could be misread in terms of RAIB’s current operations.” Indeed, RAIB’s performance is rather worse now. “…if made public, the 2007 data would generate further correspondence that would have to be addressed. This would divert RAIB’s limited resources away from the primary activity of investigation and publication of its findings.” Classic! “Analysis of the survey data needs a degree of expertise to interpret…correctly.” Deliciously patronising! “Release of the data would provide potential for misuse and alternative emphasis or misinterpretations…” In other words, the truth might emerge. There was a whole page of this stuff.

Interrupted by bursts of laughter, I addressed each of these points methodically and appealed against the decision, triggering an internal DfT review. This was carried out by a senior official who passed judgement on the two conflicting positions in a fair and thorough manner. He too could not support most of Ms Griffiths’ arguments but did agree with her on the issue of confidentiality. As a result, I have since received an eight-page document including a lengthier summary of the report’s findings and two additional pages of extracts. (None of these have appeared in the Freedom of Information area on the DfT’s website, contrary to my expectations.)   

It says much about the Branch that, when presented with a valid and entirely auditable concern, its first instinct was to perpetuate bluff and bluster, justifying its routine failings by highlighting the wholly exceptional accident at Grayrigg. It would have been better served by accepting that regulations are not ‘expections’ and that repeatedly failing to comply with a generous one is not sustainable. Every organisation has a duty to improve, particularly those funded from the public purse. A specific undertaking to put in place actions to shorten inquiry timescales would have been the rational course of action; instead I was met with a stonewall which Adonis later helped to repoint.

RAIB’s organisational structure includes four principal inspectors together with a further 22 regular inspectors - an investigative team totalling 26 members. When they were recruited in 2005, the most impoverished of them was welcoming £55,000 annually into their bank account. In the 12 months to 30th November 2008, RAIB launched 29 inquiries. So if each of its inspectors had since completed just one inquiry, the Branch would boast a 90% compliance rate with Regulation 11(2)(a). Sadly, it doesn't come close to that.

So what were the full survey findings from which part of its 2007 Annual Report was distilled? Were they really half-heartedly upbeat about inquiry timescales? Judge for yourself. Let’s first revisit that extract from the Annual Report -

“Stakeholders accept that RAIB seems to publish reports quicker than other safety organisations in the past, but would still like to see RAIB reports published earlier. In saying this, however, they may be overlooking the time required for comments, the mandatory consultation period and subsequent consideration that has to be built in. This can add several weeks to the publication process.”

Now here’s the relevant section from the survey -

“The RAIB is rated poorly, however, on the length of time it takes to produce final reports with only 42% agreeing that the length of time is reasonable. This middle management view is heavily backed up by senior stakeholders who agree that the timeframes are just too long - although in fairness RAIB seems to be quicker than other safety organisations were in the past. Stakeholders tend to think that a more reasonable timeframe would be around six months. In saying this, however, they may be overlooking the time required for comments, the mandatory consultation period and subsequent consideration that has to be built in. This can add several weeks to the publication process.”