I know lots of things. I know that 935 Iranians applied for asylum in Britain in the third quarter of this year and I know that the price of pigs in the UK, France and Germany followed a very similar cyclical pattern throughout the 1990s. I know that there were 134,557 recorded crimes in Sussex in 2003-04 and I know that my son's primary school had an unauthorised absence rate of 1.1 per cent last year. I know these things because the government collects and publishes all these facts and millions more.
I also know that there are 15 officials on the Ministry of Defence press desk in Whitehall. So I called one of them to ask about casualties in Iraq. First, I asked how many British fatalities there were during Operation Telic. The answer is 74. The youngest was 18, the oldest 55. Each one is recorded on the MoD website, which shows photos and biographies and best wishes to the families. The site is decent and dignified, and so it should be.
Next, I asked if the MoD knew how many Iraqi civilians had been killed in the conflict. "No. There is no definitive figure on that."
Roughly how many? "We don't have a figure. The government has not tried to put a figure on it."
A thousand? Ten thousand? A hundred thousand? "No."
Are you counting? "We haven't tried."
The MoD knows exactly when and where every British soldier died but it cannot estimate the number of dead Iraqi civilians to the nearest ten thousand.
A military source at the MoD told me the army could get at least a rough estimate of casualties from the daily contact reports. He said: "We don't keep a tally because no one really wants to know. Downing Street's worried that a big number could cast doubt on the legitimacy of the war. It's a question of proportionality."
For a war to be "just" and for the use of force to be legitimate, it must be "proportionate" to the expected military advantage. The Geneva Convention does not stop armies from killing civilians, it just says they must try really, really hard not to kill more than they have to. Too many innocent deaths implies a disproportionate use of force and that is illegal.
In late October, the medical journal the Lancet published a report saying that about 98,000 Iraqi civilians had probably died as a result of the invasion. A web-based group called Iraq Body Count (IBC) collates reports from media outlets and official sources, and gives a figure of (at the time of press) between 14,571 and 16,750 civilian deaths. It does not differentiate between those killed by foreigners or by Iraqis but says its total is an underestimate and that many more have gone uncounted, partly because it is difficult for journalists and other observers to travel around Iraq. At first, the Foreign Secretary, Jack Straw, rejected the IBC figures. Then, faced with the greater death toll suggested in the Lancet, he used the IBC data to cast doubt on the higher figure. Now he asks us to ignore the IBC again.
I, too, have doubts about the 98,000 deaths. Over the past 18 months, that would work out at around 180 deaths each day and, I suspect, there would be more obvious evidence of such carnage. But these are "doubts" and "suspicions". I cannot "know" because the British government does not count the Iraqi dead, even though it counts almost everything else in public life.
In 1999, Tony Blair wrote: "I believe that having access to official statistics which we can all trust is essential in any healthy society. Statistics encourage debate, inform decision-making, both inside and outside government, and allow people to judge whether the government is delivering on its promises." Now, there's an idea.