Afghanistan Analysts Network – English

War and Peace

Counting victories and losses: the war of stats

Kate Clark 5 min

In the last week, there has been a lot of discussion about what sort and how many Taleban are getting killed and captured, how many attacks the Taleban are launching and who, indeed, is ‘winning’ the war. Is ISAF beating the Taleban back or has it been exaggerating its claims? Kate Clark has been trying to unpack some of the statistics, looking at how discrepancies may arise because of differences in who counts what and whether attacks on civilian as well as military targets are included.

Attacks initiated by Taleban and other insurgent groups are down by 8 per cent this year (January to September 2011, compared with January to September 2010), says ISAF (click here). Indeed,  if you compare July-September this year with the same period in 2010, it says, they are down by more than a quarter.* ISAF has been briefing hard on the figures. They are proof, it says, that the Coalition has not only halted the momentum of the insurgency, but is now seeing its ‘reversal’ and ‘regression’.**

Attacks by Taleban and other insurgent groups are up by 24 per cent for July-September of 2011, compared with the same period of 2010, says the Afghanistan NGO Safety Office (ANSO) in its regular quarterly report. ANSO has recorded ever increasing numbers of attacks, year on year, using the same methodology since January 2006, and is confident with its assessment that, ‘the opposition retains the strategic momentum and is increasingly tactically competent’ (emphasis in original).

The average monthly number of ‘security incidents’ rose by 39 per cent in the first eight months of the year, compared with the same period in 2010, the UN Secretary General said, in his September  report on Afghanistan to the Security Council and General Assembly (click here).

Statistics matter, especially when they make the difference between (looking like you are) winning the war or being on the back foot. In the past, ISAF, ANSO and the UN (UNDSS) statistics on the level of violence generally followed the same trend – although the absolute numbers varied. ISAF is now out of step with both ANSO and UNDSS. It reckons a down turn in attacks began in May of this year; the change was discussed for the first time when General Petraeus left Afghanistan (for an assessment, see here).

Where might the discrepancy come from? Both ANSO and ISAF count only what they call ‘kinetic’ attacks (i.e. active, military incidents) initiated by the Taleban and other insurgents, including those using small arms, mortars, rockets and IEDs. ANSO counts all attacks, regardless of the target, whereas ISAF only counts those which target their own international forces and  attacks on joint forces, including Afghan-led ones.

UNDSS tracks a much broader range of incidents  – anything which affects the security of UN staff and humanitarian workers.  This pulls together not just military engagements, but also night letters, kidnaps, the discovery of unexploded IEDs and so on.

Speaking to senior officers in ISAF, their belief that the surge is starting to look successful appears genuine. They feel they are throwing everything they have at the Taleban, that difficult fighting is bringing results, and that other parties are unaccountably failing to acknowledge this.

However, there may be a different explanation for the diverging trends of violence – if the Taleban are increasingly attacking Afghan National Security Forces (ANSF) and civilians – including Afghan civilian government targets, rather than the international military. The most recent UNAMA report on the protection of civilians (downloadable here) showed a sharp increase in the numbers of civilians killed by the Taleban, both deliberately, eg by assassination, or through recklessness, for example, IEDs which are supposed to target the Afghan or international military, but frequently end up killing civilians. Indeed, IEDs cause almost a third of all civilian deaths by all parties in the conflict.***

What this might mean for the statistics is that if an IED explodes under an international military vehicle or when a foreign soldier treads on it, that would be counted as an attack by ISAF, ANSO and the UN; if it hit a car full of civilians, or a child walking to school, ANSO and the UN would still count it.  If it was found unexploded, only the UN would include it in their statistics.

If civilians are starting to bear more of the brunt of Taleban attacks, on the ground this hardly looks like winning. In a particularly bleak assessment by the International Committee of the Red Cross (read here), its head of operations for South Asia, Jacques de Maio, described the security situation in many areas of the country as ‘alarming’, saying that many rural communities feel vulnerable as never before. ‘In many areas, including in the central regions of Wardak and Logar,’ he said, ‘[people] say they no longer feel safe because they are being intimidated and coerced by all parties into taking sides.  All they actually want to do is to keep out of harm’s way.’

Moreover, as the international armies start to withdraw in larger numbers and increasingly – as ISAF and the Afghan government hope – Afghan forces take over more patrolling and fighting by themselves – ISAF’s method of counting will inevitably lead to reductions in what they count as ‘enemy-imitated attacks’.  Expect more claims of progress, then, whether or not the Taleban stay as active as ever.


*On ANSO, ISAF and DSS graphs, there is an annual bell curve showing the frequency of  attacks/security incidents which peaks in the summer and falls back in the winter.  ISAF’s graph shows this bell curve peaking at a lower level this year  – although still higher than the peak in 2009.  The ANSO and DSS graphs show an annually fluctuating, but inexorably climbing, line.

One possible partial explanation for the ISAF’s 26% reduction in July-September  2011, compared with 2010, is that insurgent activity has peaked in different months within the overall bell curve, depending on other events. In 2009, the peak came with the presidential elections in August and this year, in the pre-Ramadan month of July.  Last year, there is a discrepancy in which was the peak month – ANSO and DSS found the peak of incidents in September – when parliamentary elections were held, ISAF in August.  In other words, ISAF’s selection of July to September as the period under observation may not be comparing like with like.

** See for example, this quote in a report in the New York Times, from a senior, un-named Coalition official talking about the momentum of the insurgency (find it here): ‘Before we had arrested it, which means to that point, they weren’t making that forward progress… What we’re seeing now is a reversal, which is, they’re regressing at this point.’

Another interesting attempt to assess progress can be found in the often war-sceptical Guardian. Richard Norton-Taylor and Nick Hopkins point to a dissonance between an optimistic UK military which is encouraged by spectacular reductions in attacks in the central districts of Helmand where British marines have been posted and a more pessimistic Foreign Office which sees a ‘bleaker’ broader picture. Norton-Taylor and Hopkins write that, ‘Helmand may be much more secure than it was, but it represents a small fraction of the country. And even the British characterise the progress there as ‘fragile and reversible.’ (click here)

*** The Taleban break the laws of conflict by directly targeting civilians, failing to distinguish between military and civilian targets and failing to take precautions to protect civilians.


Taleban ISAF