In April I wrote an article Clear trends in Iraqi violence which exposed some of the myths in the mainstream media about Iraq. With hindsight it could be argued that I chose the best possible time to write the article. February and March showed the lowest levels of violence for six to nine months. The trends were indeed clear, indeed the number of US troop fatalities had fallen every month for six months and stood at one third of the autumn level.
Almost immediately after I wrote the article, most of the indicators rose. April itself saw the biggest jump in US troop fatalities in over a year, though not reaching the absolute levels of six months earlier, let alone the spring and autumn of 2004.
April does not seem to have been a fluke. Though most of the indicators have fallen since then, most (including US troop fatalities) are still above the level of March. The numbers of Iraqi security personnel killed were also well up in April on the March and February figures, but in May and June fell back below, to the lowest recorded levels.
One of the key indicators - the one most widely quoted in the west - Iraqi civilians killed by acts of war, has shown almost exactly the reverse trend: rising each month this year, EXCEPT April. (But again, still far lower than during the summer of last year.
While the figures available in April showed a clear pattern of improvement, today's figures are far more confused. Is there a narrative that explains the erratic recent figures consistent with the previous improvements? I think, perhaps, there is, and I offer it as a tentative hypothesis.
The six to nine months to March showed improvements because the jihadists - a coalition of Iraqi baathists and foreign al Qaeda fighters - had failed in their objective of preventing the elections, and were engaged in internal bickering. As a result they were both suffering from degraded ability to strike at either the Iraqi government, or the coalition, and an internal focus. There was a great deal of talk at that time of Jordanian al Qaeda leader Abu Musab al Zarqawi being pushed aside and replaced by an Iraqi.
But bin Laden and Zarqawi reasserted their control of the jihad. Both issued video statements and there was a temporary uptick in the ability of the jihad to hit security forces. It did not last. Acting on intelligence - almost certainly from rival jihadist factions - the coalition caught up with Zarqawi and killed him. The tension between baathists and al Qaeda has now reasserted itself as rival leaders seek to fill the vacuum.
So why is it that while Iraqi and coalition forces seem safer than at almost any other time Iraqi civilians are still being killed? I suspect the answer lies in the fact that terrorists are, technically, civilians. They don't fight in military uniforms and their identities are often unknown. As the power vacuum in the jihad widens, rival terrorist factions are killing each other.
In the short term, this negatively impacts the statistics. Terrorist deaths are added to those of the genuine civilians and cause an increase in ‘civilian’ deaths.
But in the long run, the ability of the terrorists to disrupt Iraq and attack security forces is being continually degraded. Partly, they kill each other. When that proves impossible, they will continue to leak information to the government or the coalition, so that the legal security forces can wipe out their rivals for them. It is a bloody and chaotic way for an 'insurgency' to collapse. But if I am right, collapse is exactly what we are seeing.