Author’s Note: This was supposed to be about 1,000 words and somehow morphed into a paper-length monstrosity, which means some of the key points are going to get lost in some maths and charts and other things that I treasure like Smeagol did the ring. Thanks to some positive feedback along that line, the following deluge of words can be summed up thusly:
The number of untrained personnel in the Afghan National Police (ANP) is increasing, not decreasing, in an attempt by the International Security Assistance Force (ISAF) to put untrained recruits into the total force. This is being done in order to make the “end strength” number look better than it actually is. Even though it’s doing that, ISAF’s own recruiting/retention/attrition numbers show that the ANP is going to start losing personnel, not gaining them, within the next few months.
As supporting arguments, I offer the following:
- The number of untrained ANP is increasing. Not decreasing. Between October of 2011 and March of 2012, the percentage of untrained personnel in the Afghan Uniformed Police (the largest division of the ANP) increased from 15% to 20%. Which makes no sense, unless NATO Training Mission – Afghanistan (NTM-A) (the ISAF subordinate command responsible for training the ANP) is putting untrained recruits directly into the force without training them. This is a separate issue from the attrition conversation, but it’s related to the fact that ISAF isn’t being terribly transparent in its numbers reporting.
- Using ISAF’s own numbers, annual attrition is around 16% of the entire Afghan National Police (ANP) force. Attrition is best defined as “people who just stopped showing up for work.” ISAF’s numbers do show a downward trend in attrition, which is good. It means that fewer people are wandering off every month.
- Recruit numbers are also showing a downward trend over time. This is bad. It means that fewer people are showing up to be ANP.
- Retention numbers are also showing a downward trend over time. This is also bad. “Retention” is military/ISAF-speak for “people finishing up their contract and then signing up again.” Not only are people not showing up for work (attrition), they’re not getting new people (recruits), and the people they do have aren’t sticking around (retention).
- Recruits + Retained Personnel – Attrition = Total Force. The ANP is going to start cutting into the existing force over the next few months. That’s because more people are being lost due to attrition than are being replaced through recruiting and retention.
- ISAF keeps reporting increases in total numbers. Every month. That does not make any sense when calculating troop strength using ISAF’s own numbers for recruiting, retention, and attrition.
The graph for the “total force” formula looks like the below. Note that this data is for the last two years, using ISAF’s own ANP figures.
Other than that, NTM-A is doing great, and the ANP are going to be just fine.
Now for some maths.
Reading ISAF numbers about anything that could be a measurable metric for progress always warms the cockles of my blogging heart, so an event like this promises all kinds of ISAF-y goodness.
ISAF media roundtable w/BG Kelly Thomas, dep cdr for police NTM-A , on AFG police development. Begins now.
— ISAF (@ISAFmedia) July 5, 2012
Before I go further, a brief explanation of abbreviations that will be used throughout this post is in order.
NTM-A: NATO Training Mission (Afghanistan) is the organization that, in 2009, took over Afghan National Security Forces (ANSF) training, to include the Afghan National Police (ANP). The ANP is an arm of the Ministry of the Interior (MOI) and is comprised of three primary groups:
- AUP (Afghan Uniformed Police): this is the largest arm of the ANP. These are the guys with the crazy white hats waving their LEGO paddles futilely at Kabul traffic when they’re not writing fake tickets or getting high.
- ANCOP (Afghan National Civil Order Police): riot cops, essentially. Not as plentiful as the AUP, and the group with the highest attrition rate.
- ABP (Afghan Border Police): also known in Kandahar as Abdul Raziq’s private army, they’re responsible for border enforcement. In Raziq’s case, this means making sure that the drugs make it through OK.
There are several other special police units, but their numbers aren’t significant enough to mention here. Plus, in the regular reports to Congress, ISAF/NATO only references those in the hushed tones one always must use in order to explain anything related to special operations units, who are responsible for their training and mentoring.
Also, I’ll be referring to the “1230 reports” throughout this post. Those are regular reports prepared for Congress as required under sections 1230-1231 of the National Defense Authorization Act (NDAA) for fiscal year 2008. I’d go into more detail, but it would be boring. Basically, twice a year (or so), the Department of Defense (DoD) commissions these reports (at a cost well into the six figures every time) to show Congress how well the war in Afghanistan is going. Occasionally they’ll be reported on by the press, but let’s face it, there’s Kardashians doing stuff, so the incentive to dig into these reports deeply really isn’t there. I, however, have free time and apparently very little in the way of imagination, so I read them. They help me alternately sleep well or cry myself to sleep, depending on the day. However, for all their dryness and mastery of the copy-and-paste, they are interesting in that they’re the only document on progress that the DoD releases consistently.
Years are so overrated.
To continue along the original thread, the ANP play a key role in the future stability of Afghanistan, so a statement like this is interesting:
Thomas: Last 9 months, police had only 1.4% attrition rate. Very much in line w/goals
— ISAF (@ISAFmedia) July 5, 2012
1.4% attrition? Over 9 months? That’s…really pretty impressive. Only 1.4% of all ANP personnel are being lost due to whatever they’re calling “attrition.”
Unanticipated losses, included Dropped from Rolls, Killed in Action, Deaths (non-combat), Captured and Permanently Disabled (Exempted). Soldiers and NCOs AWOL more than 45 days and officers AWOL more than 30 days are Dropped from Rolls, which composes the majority (over 90 percent) of attrition.
Ah. So that’s attrition. At least that’s how they defined it in 2011, and over 90% of monthly attrition numbers are due to people just not showing up for work anymore. But only 1.4%? That is all kinds of awesome. Except for the fact that that’s on a monthly basis. Over the last 12 months (well, last 12 months that ISAF has published figures…from March 2011 to March 2012), when comparing the data in the 1230 reports, it’s actually closer to 17% per annum.
But good news, attrition is actually on the decrease, and graphically, attrition looks like this, with the blue line representing what percentage of that month’s total force was lost each month due to attrition, and the black line representing the attrition trend between March of 2010 and March of 2012:
So that’s…a good thing. Seriously, reduced attrition is good. However, the 1.4% applies to the total force, which is comprised of the various policing arms described above (ANCOP, AUP, ABP, PRC, AARP, ACLU, and the YMCA). Specific arms of the whole group have much worse problems with attrition. For example, in October of 2011 the DoD reported this:
Although overall attrition in the ANP has remained near target levels for the past year, high attrition continues to challenge the ANCOP in particular, which has experienced an annual attrition rate of 33.8 percent; although this has decreased significantly from 120 percent annual rate in November of 2009, it remains above the accepted rate for long-term sustainment of the force.
Fortunately, by April of 2012 the ANCOP had resolved that particular issue.
Though the ANCOP still suffers from significant attrition levels, averaging 1.9 percent over the past six months, the ANCOP continues to meet growth objectives.
Again, that 1.9% sounds all right, but it adds up to 11.4% over a 6 month period, and 22.8% over the course of a year if the trend stayed consistent. So what the DoD refers to in its reports as “the premier police force in Afghanistan” has the highest rate of attrition in the entire ANP. However, keep in mind that 1.4% across the entire force is acceptable, so 1.9% isn’t all that bad.
Yeah, we thought about keeping track, but then Kardashians.
Now for some numbers.
Thomas: Training goal: 157k police by Nov. Now at about 147k. Will continue to professionalize those 157k through 2014
— ISAF (@ISAFmedia) July 5, 2012
This is where it gets a little perplexing. As of March 2012, ISAF reported that there were currently 149,642 members of the ANP. That was the actual number reported, while the goal was 147,943, or, “about 147K.” Now I know that a) this is a tweet, and so b) I shouldn’t mess with it, but c) this is the 21st century, so tweeting on behalf of an organization means you get held to something resembling facts. The discrepancy means one of three things:
- There’s been a 2K loss of personnel in the ANP since March. That’s going to put a crimp in their projected December 2012 end-strength of 157K . Troubling.
- NTM-A only updates its numbers twice a year for the 1230 reports and so doesn’t keep track more than every six months. More troubling.
- BG Thomas has no idea how many ANP there are, and is working off of old information. Most troubling.
I don’t know which of these is true, but for reference, here’s my source for those numbers, taken from the April 2012 1230 report.
It’s not the first time that NTM-A has had trouble keeping track of the number of ANP, however. In October of 2011 they made this discovery:
Despite indicating positive developments in ANP force generation, NTM-A recently determined that 3,940 officers and 6,733 patrolmen were filling NCO billets; large numbers of officers and patrolmen placed against vacant NCO positions overstates the development of the NCO ranks.
Keeping in mind that this was two years after NTM-A assumed the police training mission. It took them a full 24 months to figure out what billets the police were filling in the first place. However, it’s apparent that NTM-A has bigger problems with its ANP reporting than just knowing how many ANP there are. This tweet, for example, offers some interesting dilemmas.
@watandost Very true. This is why the numbers goal is set for 11/12, which gives ANP until 2014 to become a very professional force
— ISAF (@ISAFmedia) July 5, 2012
While it’s true that a) the ANP are meeting their overall recruiting goals, so they b) seem to be on track to meet the goal of 157,000 personnel in the force, there are a couple of other factors that need to be considered and which are, in and of themselves, even more disconcerting.
The first figure is monthly recruiting, which is essential in order to maintain continued expansion of the force. This chart shows monthly recruiting as a percentage of each month’s total force. The blue line has been derived from the numbers ISAF used (see above), and the black line depicts the trend for those numbers. This shows what percentage of the total force each month are recruits. I’ll explain later why doing that presents some problems for the future of the ANP, but even if that isn’t considered in the total force each month (which it shouldn’t be, but I’m not sure that it isn’t, since the DoD is a little obtuse with its numbers in the first place), this is still a problem:
I’m not a statistician, but I’d venture the assumption that overall recruit numbers are dwindling. Even if the trend just shows straight numbers, and not percentages, it looks like this.
Which isn’t nearly as bad as this next one showing retention percentage each month. Again, the blue line shows the percentage of the total number of ANP personnel that were retained each month. Retention is re-enlistment…in other words, the percentage of the total force that, each month, signed up to stay in the ANP past their original time of commitment.
Notice that this percentage never gets close to one percent of the total force. This means that the ANP aren’t able to keep the recruits they already have. This is better shown as a percentage, because I can state with a fair degree of certainty that those being retained were already part of the force. This lack of retention is problematic for a few reasons:
- It makes it harder to achieve your desired goals when it comes to numbers.
- People aren’t seeing service in the ANP as a viable career option. For whatever reason, they’re signing up for the initial hitch, then leaving.
- This kind of turnover is going to result in a force that, while it’s “trained” on paper, is lacking in overall experience, as consistently less than 1% of the force is being retained. In fact, over a two year period, from March 2010 to March 2012, in total only 1.14% of the force was retained.
So let’s talk some net trends. This next chart shows net gains using the following formula:
(Rc %) +(Rt %) – (At %) = Net percentage gain in the total force
In this equation I’ve used the following:
- Rc = % of the total force that is new recruits
- Rt = % of the total force that are being retained (re-enlisting)
- At = % of the total force being lost each month to attrition.
Adding the percentage of recruits to the percentage of retained ANP, then subtracting the percentage for attrition, the trend is distinctly negative, and within a few months, if current trends continue, the ANP will start missing its manning goals. Consistently. However, since I’m not sure that the DoD/ISAF is counting recruits each month as part of the total force, I’ve applied the same math, but just using numbers, since recruits would be part of that equation. That chart looks like this:
Keeping in mind that ISAF has only reported numbers through March of 2012, and the above chart was built past that point using a trend function in Excel. Also it’s key to understand that not all recruits will make it through training, which makes those projections even more negative. Unless, of course, you’re just assigning recruits without training them. Regardless, the chart above only projects that net gains per month will fall below zero at a certain point..even up through March of this year, recruiting and retention are still trending downward, and if that trend continues, the ANP will not reach its goals by the end of this year. I know there will be differences in actual numbers, since there can always be fluctuations in actual vs. projected figures each month, but here are some facts:
- Recruit numbers are fewer each month.
- Retention numbers are fewer each month.
- Attrition is trending downward, but at a slower pace than recruits and retention
- This is not a sustainable model as the necessary inputs are in decline.
Even though recruiting and retention percentages are dwindling, the following shows the monthly trend for NTM-A’s ANP numbers.
Math is hard. Let’s do less math and make more charts.
Which makes the whole thing perplexing. The NTM-A figures show that recruiting and retention numbers are in no way going to be able to increase the force at the rate necessary to reach the end strength goal of 157,000 personnel by the end of the year, but the end strength numbers NTM-A is reporting for each month are consistently high enough that they’re going to reach their goals without any problems whatsoever. This same problem appears when applying basic math to NTM-A’s numbers month-to-month. For example, using August and September of 2010 to illustrate my point:
A couple points of explanation:
- The negative in the “Difference b/t Goal, Actual” numbers means the goal was exceeded. For example, in August they ended up being at 102.56% of their projected goal for the month.
- This changes in the “Difference b/t Attrition, Recruitment” row. In that case, the “-42″ on the right means that more people were lost due to attrition than were recruited that month.
To break this down, then:
In August there were 110, 087 members of the ANP.
In September, there were 1767 recruits.
In September, there were 113 ANP members retained.
In September, there was a loss of 1809 ANP members due to attrition.
110,587 + 1767 + 113 – 1809 = 71
However, by the end of September, there are now 110,559 members of the ANP. Somehow, the ANP gained 472 personnel in that 30 day period. Granted, technically one shouldn’t count recruits, since technically one should not count untrained personnel as part of one’s total force. That’s going to be important later. Which means that only 472 personnel completed training in the month of September. If that were the case, NTM-A would only be able to train 5,564 personnel over the course of a year.
More math. It doesn’t get more pleasant.
So the numbers aren’t adding up. This kind of numbers shuffle happens in these reports from time to time, and is usually caught right after the last report came out and right at the beginning of the six month reporting period for the current report. For example, in the April 2011 1230 report:
In late October 2010, NTM-A/CSTC-A discovered a procedural discrepancy in the personnel accounting process for ANP resulting in some double-counting. As a result, the end-strength numbers for August-October 2010 were restated. NTM-A/CSTC-A subsequently sponsored an assessment team that validated current accounting processes and corrective actions. They validated both the ANP end-strength and the accuracy of the November and December 2010strengths. As of March 31, 2011, the ANP reached an end-strength of 125,589, exceeding by 3,589 the expected March position of 122,000.
So right after they submitted the report for the period ending in October, they discovered they’d made a “mistake.” Those kinds of number resets will set one back. Also, not having as many active training sites as one had planned will have similar results. Also from the April 2011 1230 report:
Overall, ANP training capacity is evolving to meet current and future ANP training requirements. NTM-A/CSTC-A is continuing to expand training sites in order to grow the size of the ANP and develop the capacity to professionalize the force. Between October 2010 and March 2011, training capacity increased from 11,252 to nearly 15,000, and will exceed 19,000 by December 2011.
As is often the case, however, the 1230 report makes a change in numbers without explaining why that change occurred in the first place, as is true in the October 2011 report:
Between April 1, 2011 and September 30, 2011, ANP training capacity increased from 12,822 to nearly 14,500, and is expected to reach approximately 16,000 by the end of December 2011. NTM-A/CSTC-A continues to seek efficiencies while developing the necessary capacity to grow the size of the ANP and professionalize the force. Training is currently conducted at 32 formal training sites, but this total will eventually decrease to approximately 12 sites in 2014 as the permanent training base is established and temporary sites are closed. Across all police pillars and all courses, 18,016 students have graduated since the beginning of April 2011.
I have no idea how an organization that can only train 16,000 personnel graduated 18,016 students: my guess here is that it’s due to the fact that training is not being completely executed by the ANP, but is still being done by ISAF at some sites. If that’s the case, then around 4,000 are being directly trained by ISAF, with the remaining 14,000 trained by the ANP.
There’s also no explanation given how in six months the projected numbers went from exceeding 19,000 to approximately 16,000. That explanation doesn’t appear for another six months in the April 1230 report:
Between October 1, 2011 and March 31, 2012, ANP training capacity increased from nearly 14,500 to 14,584. The ANP was expected to reach approximately 16,000 personnel by the end of December 2011; however, severe delays at National Police Training Center (NPTC) – Wardak impeded achievement of this goal.
So in the course of a year, from April 2011 to April 2012, instead of the projected training capacity reaching 19,000, and an increase of 4,000 personnel, the training capacity only increased by…84 personnel. That’s assuming that “nearly 15,000″ in April of 2011 is the “14,500″ that was listed in October 2011 and again in April 2012.
To review up to this point: none of the numbers that NTM-A is releasing to Congress and to the public via the 1230 reports appear to be adding up. The projected training capacity is lacking by several thousand personnel, and there is no way possible, based on the calculations above, that NTM-A will meet its force projections by the end of 2012 based on the current trends in recruiting and retention.
It gets worse.
Unfortunately, that’s not the really bad news. In the April 2010 1230 report, ISAF/NATO/NTM-A realized that the ANP was using what they called the “recruit-assign” model. In this particular model, a new recruit to the ANP would be assigned to a unit without training, so they would be counted against the ANP’s total end-strength, but as an untrained member of the force.
Training is a key challenge to building the capacity of the ANP. In recent years, because of the lack of program resourcing, 60-70% of the force was hired and deployed with no formal training (the “recruit-assign” model). While working to increase the throughput of new recruits, NTM-A/CSTC-A has also had to implement plans, in coordination with the IJC and the MoI, to provide training to those already on the tashkil. For basic recruit training, one of the major initiatives, implemented in March 2010, is the establishment of a “recruit-train-assign” model. “Recruit-train-assign” will ensure all new police recruits receive necessary training before performing official duties.
Fair enough. One realizes one has a problem. One addresses that problem. One makes sure that one a) goes back and trains those already untrained, and b) going forward, that one is not assigning untrained personnel to police units. Just as a point of information, this issue of untrained ANP was first mentioned in a January 2009 progress report, but was not referenced again until over a year later. This isn’t a problem that’s going to resolve itself quickly, so this from the October 2011 1230 report isn’t terribly surprising:
The Afghan Uniform Police (AUP) is the largest of the Afghan police pillars; as of September 2011, the AUP end strength was 79,432 personnel. The AUP is on schedule to meet all growth objectives for officers and patrolmen, and is capable of achieving 98 percent of its 2011 NCO objectives under current training fill rates.
Untrained patrolmen remain the biggest challenge for the AUP, and NTM-A/CSTC-A and the
MoI continue to push the recruiting base in order to ensure all available training seats are used. As of September 2011, the AUP had a total of 11,919 untrained patrolmen and NCOs.
So 15% of the total force was still untrained. That’s down from 60%-70%, and is a number so positive as to be nearly unbelievable. That point is illustrated by the April 2012 1230 report:
As of March 2012, the total strength for the AUP was 85,434, an increase of 6,002 personnel from the previous reporting period. The AUP is slated to grow to an end-strength of 85,532 personnel by November 2012.
Untrained patrolmen and the lack of a sustainable logistics system remain the biggest challenges for the AUP. NTM-A and the MoI continue to emphasize recruiting in order to ensure all available training seats are used. As of March 2012, the AUP had a total of 12,500 (20 percent) untrained patrolmen and NCOs. AUP attrition remains the lowest of all police pillars, averaging 1.0 percent per month during the reporting period.
In the course of six months, the number of untrained patrolmen and NCOs in the AUP has now increased by 5%, but if NTM-A is using the “recruit-train-assign” model, the numbers should be going the other direction.
Which leads me to believe that the “end strength” numbers being reported each month actually include recruits who are not being trained, but are being assigned without training in order for ISAF and the MOI to be able to report monthly increases in available personnel. Which, is disconcerting enough, given how important it is that the ANP be as professional a force as possible between now and the inevitable drawdown in 2014. What’s more disconcerting is the fact that ISAF isn’t reporting similar training concerns for the other branches of the ANP. Given the shortcomings they have reported (high rates of attrition, inability to determine what rank is slotted in what billet), I’m concerned about what they’re not reporting.
Let me be clear on one thing: having spent more time than I care to admit making graphs and doing math, nothing would make me happier than for someone to show me how a) my math is all wrong, b) the numbers really do add up, and c) that things are going to be just fine.
But some things lead me to believe that isn’t the case:
- The definite downward trends in retention and recruiting.
- The lack of transparency on training capacity shortcomings.
- The fact that the reports don’t define “end-strength”: does it include trainees? How many are actually trained in the force?
- The fact that there are more untrained AUP patrolmen in March of 2012 than there were in October of 2011.
It’s that last one that I find most troubling. That is irrefutable, simple math based on the 1230 data itself. The numbers are there, and the only logical explanation for an increase in that figure is that NTM-A/MOI is still using the “recruit-assign” model.
I’m not even dealing with the other information about the ANP: the fact that to increase ANCOP recruitment NTM-A lowered initial entry literacy standards, or that the majority of the ANP is still illiterate, or the issues with pay, or the difficulties in establishing logistics support. I’ll likely cover those topics at another time.
What I am dealing with is the incontrovertible fact that ISAF (and in this case specifically NTM-A) demonstrates a distinct lack of transparency in its reporting. Figures like the “1.4%” are consistently released in such a way that it would be easy to conclude that this figure covers multiple months. While presenting the appearance of transparency by making accessible to the public these bi-annual reports, the United States Department of Defense, ISAF, and NTM-A have yet to publish any report that is truly transparent in its reporting or methodology.
This will not be the last post I will do on this topic because it’s important that, even at my insubstantial level, someone understands what is being done here. Either no one at the DoD reviews these reports each time and therefore doesn’t see the differences being reported, or else it’s just a terribly clumsy game of three card monte.
Whatever their motivations, what is abundantly clear is that the current training situation for the ANP is untenable, unsustainable, and will ultimately be unsuccessful. Playing with the numbers only hides the fact that while the future of the country may be uncertain, the future of the ANP is not. And it is that future that looks bleakest of all once the transition, ready or not, is complete.
Until next time, you stay sunny, Afghanistan.
- Transitioning Afghanistan (dvidshub.net)
- Construction complete on Afghan National Police training facilities (dvidshub.net)
- Kabul filmmakers help reshape Afghan perceptions of ANP (dvidshub.net)