Z.04: Between Probability and Representation

Celebrating International Women's Day in Shindand, Afghanistan. Team photo, 2011.

I got ambushed my first day at ASPI, where I am a visiting US Army War College Fellow.¹ I was out of my native environment, seconded to a think tank in a foreign country and wearing a suit and tie instead of a uniform. However, I’m a graduate of Ranger School and, in the same vein as retired General Mattis’ advice, I’m always ready to execute battle drill #4.

I’d just been through a whirlwind of introductions to everyone in the office when I was invited into a roundtable with Lisa Sharland, a Senior Fellow and Director of the Protecting Civilians & Human Security program at Stimson. She gave an engaging presentation on the work the UN had done over the last 24 years as part of its Women, Peace, and Security (WPS) resolution. While I had some understanding of how the UN has worked to involve women in crisis zones in the long-term solutions, Sharland was the first person to introduce me to Gender Responsive Leadership efforts, despite being around longer than I’ve been in the army. The name doesn’t roll off the tongue but looks at sending forces and the structural barriers to females serving and succeeding in them. 

And when they opened up the room for questions, one of my new colleagues asked, ‘I’m curious what Erik’s perspective is?’ React to ambush, go!

I offered up my own experiences, citing how in 2010 during a Village Stability deployment in Afghanistan women’s issues had been something we’d worked on. I started that deployment with a USMC FET, or Female Engagement Team, which consisted of two volunteers from the Corps who helped us engage with the local community as we tried to improve not just security but also the local economy. That FET team was later replaced by a CST — Cultural Support Team — another two volunteers, this time from army SOF units. They were the ones that worked the ceremony for International Women’s Day pictured at the top of this post.

I’d even had one of the first iterations of a similar effort all the way back in 2004, when we’d attached ‘Lioness teams’ — single female volunteers from our brigade — to help with searches and cordoning the females when we did raids. Each program had its own approach to engaging with the females of a security challenge, seeing their role in providing solutions, and the costs of not taking them into account. And, contrary to the narrative that women were banned from seeing combat until 2015, each of these examples is from before then. This false narrative is more than just factually wrong, it has prevented female veterans from receiving benefits they are entitled to.

The Big Deal That Wasn’t

I do remember the first time I joined a unit with a female assigned to it — the examples above were all attachments. I’d been in the Army for over a decade, which means I spent just over half my career in units that explicitly prohibited women from any and all jobs. So when my battalion decided to allow a female intelligence officer to be the battalion intelligence officer, it was A. Big. Fucking. Deal.

Except barely six months later, it wasn’t. The editorial pages raged in much the way they did a few years earlier when we repealed ‘don’t ask, don’t tell’. Post after post decried how letting females support these units, or — god forbid, actually try out for them — would be our undoing as a military and threaten the safety of our democracy.

Meanwhile, in our battalion it was a giant nothing burger. Y2k had a bigger impact on our organization than allowing a female soldier to serve. And, in a portent of things to come, she was good at her job. Better than her male predecessor, and better than the male officer who came after her.

In my command half a decade later, top performing females were spread across our formation. Females were running the comms shop and our chow hall. We had a female company XO (with a ranger tab) and later a female headquarters company commander. I had a new standout female intelligence officer and our JAG was doing her damndest to keep me out of jail. None of these were hired to be ‘firsts’, they just applied and they were crushing it because they were incredibly talented officers and NCOs.

This isn’t to say female soldiers are better than male ones. They aren’t. There’s a bell curve of talent in every pool. But not everyone was hunting for talent across the entire formation, which meant I got access to more talent than my peers. Peers who still clung to the idea that the only talent was in the pool that looked like them. I wasn’t picking females because of their gender, but because I’m data literate, and I’ve read Tversky and Kahneman’s work on the ‘conjunction fallacy’.

Thinking Fast and Dumb

Tversky and Kahneman produced a lot of really brilliant research, which is why both Thinking Fast and Slow and Michael Lewis’ biopic of the duo, The Undoing Project, are on my suggested reading list for data literacy. Many of their key insights were about where our default thinking leads us astray. One oft-cited example is their ‘conjunction fallacy’, which goes something like this:

Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

Which is more probable?

1. Linda is a bank teller.

2. Linda is a bank teller and is active in the feminist movement.

Tversky and Kahneman’s research found the majority of respondents incorrectly chose option two. It’s wrong because statistically the odds of two conditions being true is less than just one.² So, let’s look at an inversion of the same math problem:

Jim is 42 years old, a commander of an army battalion and thinks he’s very bright. He wants to hire talented officers to join his team. In his search for talent, he comes across a resume on AIM2 for CPT Smukatela.³

Which is more probable?

1. CPT Smukaetla is a talented officer.

2. CPT Smukaetla is a talented officer and a male.

Maybe asking the question in such an overt way might help army leaders break through Tverksy’s and Kahneman’s conjunction fallacy. But too few army leaders take the time to ask the question. It’s not always overt, and not often even a conscious thought, it’s just ‘system one’ thinking when you should be engaging ‘system two’. Taversky and Khaneman described it as ‘...inconsistency between the logic of probability and the logic of representativeness.’ Or as a female officer explained it to me:

‘Overt sexism still exists, but is relatively rare. However, what I see now is that men get to come in with the presumption of competence, an advantage not afforded to women. Men have to fuck up bad to get relegated and lose trust. On the other hand, women come in with the presumption they suck – or at least don't compare to a male counterpart. Off the bat, they are presumed incompetent until they are able to prove your bias wrong.’

Leaders want talented soldiers to work in their formations, and with the advent of tools like AIM2, we have an unmatched ability to hand select the people we recruit. But only if we’re willing to look at the total pool of talent available to us. A smaller pool is most likely going to have its own bell curve of talent, with fewer overall tail-end talent. A larger pool of talent has more potential for talented officers than a smaller one.

‘Arbitrage human nature. It's not going to change any time soon.’

⁴My real secret is I wasn’t even doing anything. I wasn’t searching for talented female officers. I wasn’t trying to get any ‘firsts’, or trying to appease any leader or policy above me. I didn’t recruit a single one of these top performing officers. All I was doing was not standing in their way.

There was a whole pool of untapped talent out there, and they were applying to come to our battalions. Sua sponte. I just happened to be one of the few commanders who wasn’t screening their applications out.

Many of my peers were pricing their value at $0 dollars, so I got to enjoy the arbitrage effects on the cheap. SF battalions aren’t for everyone. The missions are challenging and the optempo is high. The soldiers who volunteer for SOF want those challenges. Among the population of non-Green Berets in our battalion, a disproportionate number were top talent. This is because higher talent people will seek out a harder job more than low-energy ones. For some of these soldiers it’s the mission. For others it’s being part of a unit that has a higher baseline of expectations, be they for fitness or job performance. The effect is a right-justified distribution of talent among our maroon berets.

I got a top tier battalion staff without having to put in real work to recruit it. I could just sit back and welcome those resumes my peers passed on. My advantage won’t last forever, but it’s been more enduring than I would have expected. In part because human nature changes so slowly. Remember, Tversky and Kahneman’s key insight wasn’t that the odds of two conditions were less than one. It was that most people get the answer wrong.

The UN’s WPS initiative celebrates its 25th year anniversary next year. They’ve made great strides, with tangible impacts across the globe, both in conflict zones and in sending forces. But there’s a lot more distance to go. Thankfully, female service members aren’t waiting on their leadership to change, but quietly making strides in every service.

As a leader, you can do more to enable females to succeed in your formations. More than just not standing in their way, you can change the way we mitigate and ameliorate the conditions that produce unequal outcomes. Things you probably never had to think about, and never anticipated are preventing you from maximizing the potential in the soldiers available to you. That’s probably going to take some self-study. I didn’t even know what Gender Responsive Leadership was until this past summer, and it wasn’t the military that taught me.

Then again, the military didn’t teach me data literacy either.⁵

1. The army accepts applications from just over 80 officers a year to complete their SSC^1.5 year abroad at various universities, corporations, and institutions in lieu of doing a year resident at the war college in Carlisle, PA.

^1.5 - Senior Service College is the educational course many colonels attend upon being promoted. Many non-military people may be surprised to learn the military places high value on education and professional development, especially compared to other professions. When I finish this fellowship, I’ll have spent just under 1/5th of my 22 years in the Army being paid to go to school.

2. For the formula inclined: For conditions A and B, Pr (A ∧ B) ≤ Pr (A) and Pr (A ∧ B) ≤ Pr (B).

3. Assignment Interactive Module 2 (AIM2) Marketplace. The online clearinghouse for jobs and soldiers for the army.

4. Quote is from Jim O’Shaugnessy.

5.  The title of this post went through a lot of revisions, and the best one, which we ultimately decided not to use, was ‘Broadening the small penis pool’.

I include that detail here as a reward for those of you who read to the very bottom.