Fun with numbers: How to fail at pushing a media bias talking point

Dafydd ab Hugh posts in an attempt to savage a recent AP article on military recruitment.  But for someone using the "mind of a mathematician", his argument falls apart.

First, the AP article states that most believe the army needs an extra 30,000 to 50,000 troops to meet its overseas commitments, and then Dafydd says:

A quick calculation:  so our force is just too small -- we desperately need to increase it... by 3% to 5%.

Already I'm scratching my head. Three percent? That's the total increase that we need to be a powerful fighting legion once more?

No...the article said that we need 30-50,000 to meet overseas commitments.  We currently have about 18,000 troops in Afghanistan and about 140,000 troops in Iraq.  That's about 160,000 in current conflicts.  An extra 30-50,000 troops in Iraq and Afghanistan would increase troops in these theatres by 19-31%. 

He then notes that the Army may miss its recruitment goals by only 8.75% and is only .7% of the entire Army.  True...but that's not the point of the article.  The article notes that this is the worst shortfall in recruiting since 1979, and this demonstrates that it would be difficult to increase the size of the Army by 30-50,000 troops.

Also, he mentions that the Army has exceeded its recruitment goals over the last four months.  Interestingly, the Wall Street Journal, in June, published an article entitled To Fill Ranks, Army Acts To Retain Even Problem Enlistees.  Some excerpts:

To keep more soldiers in the service, the Army has told battalion commanders, who typically command 800-soldier units, that they can no longer bounce soldiers from the service for poor fitness, pregnancy, alcohol and drug abuse or generally unsatisfactory performance. Typically such decisions are made at that level. Instead, the battalion commanders must send the problem soldiers' cases up to their brigade commander, who typically commands about 3,000 soldiers.

"Basically it is another set of eyes reviewing cases. It lessens the chance that we will separate people who might still make good soldiers," says Maj. Elizabeth Robbins, a Pentagon spokeswoman.

Army officials say the move isn't unprecedented. The service made a similar decision in 1998, when the strong economy and lack of a clear mission left the military struggling to meet recruitment goals.

Still, some Army battalion commanders are less than pleased with the Army's decision to try to keep more problem soldiers in the service. "It is the guys on weight control ... school no-shows, drug users, et cetera, who eat up my time and cause my hair to gray prematurely," says one Army battalion commander. "Often they have more than one of these issues simultaneously."

And some battalion commanders question whether it makes sense for brigade commanders to make decisions about which soldiers can cut it and which must go because the brigade commanders have less daily interaction with the soldiers and their immediate commanders.

One commander says the growing attrition problem can be traced to a slip in the quality of new soldiers as recruiters have increasingly struggled to hit their monthly quotas. "There are guys showing up at units with physical problems or other issues who you would not have seen a couple of years ago," says the commander.

Why brag about meeting recruitment levels recently when recently, the Army has decided to lower its standards for recruitment?

Finally, Dafydd discusses reenlistment:

But among soldiers who have actually fought in Iraq and Afghanistan, and who therefore know what is really happening on the ground from personal experience, re-enlistment rates were met. In fact, since the goal is to retain more people than normal (to meet the goal of increasing troop strength by 30,000 -- right?), it's safe to bet that the re-enlistment goals were higher than in previous years... which means there are more veteran troops in the Army than before, which would partially mitigate the lower numbers of raw recruits.

There's an important question to ask here.  What were the re-enlistment goals?  First, we have no idea what percentage of troops were expected to re-enlist to make retention, and we have no idea if these retention percentages have changed over time.  If the retention percentage goal was high, then meeting re-enlistment has some meaning.  However, if the retention percentage was either relatively low or has decreased over time, then making this goal is less impressive.

More importantly, what number of Army members re-enlisted?  I haven't been able to find anything for 2005, and (interestingly) the Army hasn't published actual retention figures in their last four recruitment and retention press releases (see the numbers for August 2005 and note that all numbers are published except for active duty retention).

The only numbers I've been able to to find for Army retention were from 2003.  In that year, the Army recruiting goal was 76,000, close to this year's goal.  What was the retention goal?  13,833.  They made their retention goal that year with 14,599 soldiers staying.

Now we have no data for 2005, and it's possible (though highly unlikely) that the goal for this year is much much larger.  But if retention goals are similar for 2005, then making retention isn't as impressive as Dafydd is making it out to be.  In fact, these numbers highlight the need for recruitment.  7,000 non-enlistees is almost half of a retention goal of 15,000.

The AP story isn't some example of liberal media bias.  They're simply telling that the Army is having problems recruiting, an important thing to know for the purpose of seeing how easy/difficult it would be to add to our active forces. 

Comments

[mike]

The article pretty clearly states that the 30-50 extra K would be on top of the million, not the 160K currently in Iraq and Afghanistan."Fraid you're wrong in your analysis.

[Eric]

Nice try at messing with things statistically, however, everyone in the Army is deployable, every unit is deployable. The new recruits will not be sent straight overseas. They will be sent to units globally. The units themselves rotate. The increase is for end strength, not combat deployed strength. The article may state that as the reason, but that doesn't make it so. Not sure about you, but I served in the military, including in Army Recruiting Command. I know a bit about how this works.

You do know that the retention figures you discuss only apply to first term soldiers who are ending their term of service that year, right? And, speaking as an ex-recruiter, I'll tell you that because of how those goals are derived (# of first term soldiers coming to end of term service), they will go up or down based on the size of the Army. With more first term soldiers (like now) those numbers went up, not down.

You do know that it requires 12 to 18 months to produce a fully competent and capable soldier, right? I'd explain how discharges and such work and how they've changed over the years, but it would take me far too long. Suffice it to say that the changes to retention of soldier who would normally be discharged early in peace time is not the big deal that you think it is.

[llamaschool]

Mike,

This is what the AP article states:

"Many in Congress believe the Army needs to get bigger - perhaps by 50,000 soldiers over its current 1 million - in order to meet its many overseas commitments, including the wars in Iraq and Afghanistan."

Yes, it states that it would be on top of the million...but that the extra 30-50,000 soliders are needed "in order to meet its many overseas commitments, including the wars in Iraq and Afghanistan", and not for anything else. The point of the article is that these 30-50,000 are important to maintain overseas commitments. To compare these numbers to the number of troops in the entire army isn't the point, because they're not necessary for the entire army. They're important towards maintaining overseas commitments.

Eric,

First, how am I "messing with things statistically"? Do you have any examples?

Yes, I do know that every unit is deployable. But you can't deploy every single unit in the U.S. Army in Iraq or Afghanistan, correct? You need soldiers at other overseas locations (Korea, Germany, etc.) and most importantly, in the States.

For example, let's say there are 160,000 troops in Iraq/Afghanistan, and 840,000 elsewhere (mostly stateside). And let's say that there needs to be 200,000 troops in Iraq/Afghanistan (which is an important point of the article).

In order to get the necessary 200,000, you need to either decrease the number of troops outside of Iraq/Afghanistan, or get new recruits. And I'm assuming that we haven't moved more troops to Iraq/Afghanistan because we need those 840,000 troops in places other than Iraq or Afghanistan. So more soliders aren't moved to Iraq/Afghanistan because this would stretch the limits of the army stateside and outside of Iraq/Afghanistan. That's why some defense analysts think it's necessary to increase the army by 30-50,000, as stated in the AP article. And this is why missing a recruiting goal by 8,000 shows the difficulty of adding troops that could help in Iraq. These aren't big percentages relative to the entire army...but they are much bigger relative to the # of troops in Iraq/Afghanistan.

As for retention numbers, it's difficult for Dafydd (and others) to say meeting a retention goal is more important than missing a recruitment goal if a) the raw retention numbers aren't published and b) we don't know the percentage of retained soldiers necessary to meet the goal. Dafydd's article doesn't have any reference to any hard retention numbers, and neither does the August 2005 press release. Without that data, Dafydd can't claim that soldiers are "re-enlisting in such huge numbers, easily exceeding retention goals".

[mike]

I'm not tryinig to get into an argument, but I think the point of the original post on these numbers (the one to which you responded) was that the coverage of recruiting "shortfalls" has been consistently alarmist, and unnecessarily so. If the current size of the Army was 100,000 men, 30-50,000 men would be huge. But the larger the base number, the less significant the 50K becomes. As for their utility in overseas commitments, well, of course. Why else would the US need more soldiers? And keep in mind that the need isn't for more soldiers actually in Iraq or Afghanistan. But again, needing a mere 3-5% increase in total numbers is hardly evidence of desperation or imminent failure. I guess another way of looking at it is to consider this: Is it realistic to think that the US might fail in Iraq for lack of 30,000 Army recruits? I think the question points out the silliness of the coverage. Finally, I'm not sure how useful it is to take what "many in Congress" believe too seriously. Who are the many? And how many is many? I don't give a hoot what my buffoon Rep thinks, for example, or what Barbara Lee thinks on this issue.

[llamaschool]

Mike,

Thanks for the response. I respectfully disagree. 50,000 troops would make a big difference in Iraq/Afghanistan, either by allowing for individual troops to spend less time in Iraq/Afghanistan, having more members of the Army in Iraq and less members of the National Guard, or by having more troops on the ground. I don't know if the Army will fail with the current number of troops, but an extra 30-50,000 would certainly help.