By the Numbers: U.S. Military Brain Drain

As the defense budget tightens and the armed forces draw down, military brass face difficult choices in reshaping the services for the future. Fundamental to force structure reductions are decisions about personnel: who exactly should comprise this smaller force?

In this environment, designating the most talented military officers and targeting them for retention seems a worthwhile—perhaps even obvious—endeavor. Yet efforts to this end are limited at least partially because military leadership disagree about whether talent attrition, or the so-called “brain drain,” is a legitimate concern. Hyperbole and anecdotal evidence often overwhelm this debate, but data suggest that the issue might be clearer than senior defense officials think.

I recently explored such data as an analyst for the National Commission on the Structure of the Air Force, a congressionally-mandated commission charged with providing strategic recommendations to Congress and the President. The Commission’s charter empowered staff to request a litany of raw data from the Air Force, including figures useful for examining talent attrition. Because of the Federal Advisory Committee Act, the unclassified data collected by the Commission, including those explored in this article, were released to the National Archives and now reside in the public domain. While not definitive, the following analysis demonstrates one framework for evaluating talent attrition trends and provides some preliminary insight.

In order to analyze these data, however, it is imperative first to explicitly define “talent.” Much of the debate on talent retention suffers from equivocation on this term. An ideal metric to quantify officer talent should include intellect, leadership, work ethic, and other criteria. But how can we capture and weigh nebulous qualities?

The solution I chose was to focus on one attribute: intellect. While this quality of course represents just one dimension of military talent, the notion that it is among the most important characteristics of an effective officer should be acceptable a priori. To approximate intellect, I charted the university from which each officer commissioned.  Since competitive universities evaluate applicants based on broad history of academic achievement including grades, test scores, and other accomplishments spanning several years, this metric should be a reasonable indicator of intellect. And if the “brain drain” is genuine, the data should show higher attrition rates among officers from more competitive programs.

To organize the data, I parsed officers into two groups: those who commissioned from “Top Ten” U.S. national universities according to the ranking of the 2013 U.S. News and World Report, and the field (i.e., all other universities). This categorization is arbitrary to an extent, as I could have just as reasonably selected “Top Five” or “Top Twenty-five.” But the study requires a partition, and since the objective is to illustrate trends as opposed to making value judgments about either group, Top Ten suffices as a division. (Undoubtedly, there are many brilliant officers in the field who make valuable contributions to the Air Force, and I do not assert via this study that intelligent officers reside exclusively in the Top Ten group). I then examined data on Air Force officers commissioned in Fiscal Years (FYs) 1995, 2000, and 2004 for both groups. I did not select these particular FYs for any reason other than to provide a range of economic and political conditions for the classes. Finally, I built cumulative attrition profiles, tracking what percentage of each class separated from active service by group for each year from commissioning (FY 1995, FY 2000, FY 2004) to the end of FY 2013.

The results are consistent with a brain drain.

Without exception, in every year after the classes had reached four years of service (the minimum amount of service required of newly commissioned officers excluding extenuating circumstances), the Top Ten group had a higher cumulative attrition rate than the field.

For the FY 1995 class, after 10 years of service, 70% of the Top Ten group had departed the active Air Force versus 50% of the field. This difference in attrition remained relatively steady for this class through FY 2013, by which time 80% of the Top Ten group had separated versus 62% of the field.

The disparity in attrition rates is even greater with the FY 2000 class. By 10 years of service, 75% of the Top Ten group had separated versus 50% of the field. By the end of FY 2013, 82% of the Top Ten group had separated versus 60% of the field for this class.

The attrition disparity was smaller but still significant for the FY 2004 class; by FY 2013, the Top 10 group declined by 62% versus 48% for the field.

A skeptic might object that the difference in attrition between these two groups is spurious because the field contains graduates of the U.S. Air Force Academy. Intuitively, Academy graduates should be predisposed to longer military careers due to self-selection – after all, these officers chose this military environment over a civilian institution, perhaps demonstrating a penchant for long-term military service. Since this factor would increase retention rates for Academy graduates, then, including these officers in the field could skew the data.

But even if Academy graduates are excluded altogether, the disparity in attrition between the field and the Top Ten group is significant. With Academy graduates removed from the data, the cumulative attrition for the field in the FY 1995 class rises just 3% after 10 years of service and 2% through FY 2013 compared with the aforementioned figures.

Similarly, for the FY 2000 class, the cumulative attrition for the field increases 2% through 10 years of service and 1% through FY 2013 with the Academy removed from the field.

Finally, for the FY 2004 class, the attrition through FY 2013 increases 4% without the Academy in the field.

So for all of the aforementioned figures, excluding Academy graduates from the data slightly increases attrition rates for the field but does not significantly change the overall attrition disparity between the field and the Top Ten groups.

Of course, this analysis is incomplete. Intellect and academic achievement alone do not necessarily indicate a talented officer, and this study excludes other important qualities of a military leader. Also, the notion that the attrition disparity indicates a talent management problem unique to the Air Force might be overstated – after all, many organizations outside of the military undoubtedly struggle to retain top performers.

Still, in tandem with other evidence like Tim Kane's survey results, these data suggest that in the Air Force, at least in the three year groups evaluated above, many of the brightest officers tend to separate at a higher rate than their peers. The services would benefit from application of this methodology to each individual career field in order to identify talent attrition trends more precisely (albeit with a more nuanced algorithm that incorporates other important metrics for military talent).

Ultimately, it is incumbent on Congress to mandate that the armed forces define talent, establish metrics to approximate this definition, evaluate retention trends against these metrics using an objective, analytical framework, and develop ways to shape those trends to deliver desired outcomes. Congress establishes limits on the quantity of officers in the armed forces; benchmarks for quality should also factor into their requirements.

The military services already excel in recruiting and developing top talent to comprise their officer ranks. It is time to address keeping this talent in uniform. 

John G. Swisher is an Operations Research Analyst in the U.S. Air Force. The views expressed are those of the author and do not necessarily reflect the official policy or position of the Department of the Air Force, the U.S. Government, or the National Commission on the Structure of the Air Force.