Yesterday, Major General Tony Thomas, who heads US Special Operations Forces in Afghanistan, fielded questions via satellite from a number of reporters gathered in Washington. The transcript of the session can be found here. One claim by Thomas that stood out to me as  I listened was an assertion that Afghan Special Operations Forces do not suffer the same high level of attrition that is seen generally for ANSF. Here is the exchange, where Thomas’ response to the first half of the question has been edited out and emphasis has been added:

 Q: General, it’s Luis Martinez with ABC News. Can I ask you about what your command’s role is going to be after 2014? What — will there be a shift in emphasis? What exactly will your operators be doing?

And also, the Afghan national army as a whole seems to suffer from attrition and retention problems. How does that manifest itself in the commando kandaks, if at all?

MAJ. GEN. THOMAS:

/snip/

We are not suffering similar attrition or, as we prefer to look at it, retention challenges with the commandos and with other formations. But again, here again, we’re lucky. We’ve been working with them for a while. They are on a cycle which has a built-in break, so it’s a great, amber, red cycle, where green, combat is on the schedule, they are going into operation and they know that they’ll have, you know, a — they’ll be applied in the hardest possible scenarios. But on the other cycles, they’ll have a chance to recoup, take leave. They’ll also have a chance to train as they come back into green cycle.

And I know that others are attempting to apply that same cycle to the rest of the force. That’s been the great challenge for the rest of the Afghan security forces, is they’re almost in a relentless combat cycle, and it’s breeding some of the retention challenges. But we are — we are looking to fix that over time, and, again, the special operations example is applicable to the rest of the force. We just need to bring that into line.

I noted at the time Thomas said this that it should be fairly easy to fact-check Thomas on his claim that Afghan Special Forces do not suffer the same high attrition rate as the rest of ANSF. One reason for my thinking this is that Afghan Special Forces are not nearly as highly trained as US Special Forces. There is only a twelve week extra training period for Afghan troops to be classified as special. I have a hard time seeing how such a short period of additional training will add significantly to retention rate.

If we look fist to Defense Department claims only, the most recent Report on Progress Toward Security and Stability in Afghanistan (pdf), dated December 2012, has this paragraph (emphasis added):

ANASOC continues to develop its institutional capacity to conduct training programs. Currently, a majority of courses taught at the Division School of Excellence are Afghan-led, with minimal Coalition Force oversight. The ANASOC has produced 11,710 Commandos and 955 ANASF. Graduation rates for both CDO and SF operators remained steady and are on schedule to meet end-strength targets. From April through September 2012, the School of Excellence produced a total of 621 new CDO, and 282 new SF operators. Based on current recruiting and graduation trends, ANASF are anticipated to achieve their end-strength force level of 1,863 personnel by the 4th quarter of 2013. ANA Commandos (ANACDO) are currently at their endstrength force level of 12,525. Staff training at all levels is occurring through uniformed and civilian mentorship programs; the target of ANASOC reaching FOC for all units is 2014, with the exception of the SMW.

There are several different categories of ANSF troops described in the paragraph, but from the context of Thomas’ remarks and the reports analyzed here, the category of commando is what is relevant. Note that this Defense Department report claims 621 new commandos trained in an approximately five month period and that the commandos are at the endstrength force level of 12,525.

However, if we check these numbers against the more independent information from the Special Inspector General for Afghanistan Reconstruction, we find both Thomas’ claim of low attrition for commandos and the current force level for commandos in the Defense Department report to be misleading.

By looking at the SIGAR quarterly reports dated April 30, 2013 (pdf), January 30, 2013 (pdf) and October 30, 2012 (pdf) we find a table in each report captioned “ANA Strength, Quarterly Change”. This table includes an entry for “Special Operations Force”. In each table, both the “Authorized” and “Assigned” troop strength levels are given for the end of the new reporting period and compared to the level at the end of the previous reporting period. There are a number of editing errors in the headings and captions of these tables, but by looking at the dates given for when data were collected and comparing over time, the following picture of ANA Special Operations Forces emerges for the most recent one year period for which data are available.

Date                                       Authorized                                          Assigned

05/21/12                                   8,224                                                   10,617

09/06/12                                 12,525                                                   10,193

11/21/12                                 12,525                                                   10,338

02/18/13                                 12,261                                                   10,366

Note that the Defense Department’s claim of Afghan Special Force size cites merely the authorized number of commandos, not the actual number that were assigned at the time, which, as of late November, 2012, was only 83% of the authorized level.

Now note that the Defense Department stated that 621 new commandos were trained from April through September of 2012. Yet, if we look at the SIGAR data, actual assigned force level dropped from 10,617 to 10,193 in the closest comparable time period (May through September of 2012). Taken together, those numbers suggest that over 1000 personnel disappeared from the ranks during this time period. If we are generous and take this loss of 1000 as representing how many are lost in six months instead of five, we still see an overall loss of 2000 commandos per year. For a force size of just over 10,000, that is an attrition rate of almost 20% a year.

Assigned force level does increase for the second half of the one year period analyzed here, but we do not have numbers on how many new commandos were trained during that part of the year, so there is no way to know if the higher assigned force level is due to greater retention or training more commandos. The only period for which we do have both numbers shows the 20% attrition. If training remained constant, then attrition would have been closer to 10% for the second half of the year analyzed and 15% for the year overall.

For the overall ANSF, the attrition rate is just over one third per year, but it seems to me that if my estimate of 15-20% turnover of commandos per year is accurate, Thomas is splitting hairs in claiming that attrition is not a large problem for them.  It is indeed lower than the overall rate for ANSF, but it is hard to see how a force can remain functional with turnover that high.