Workforce Analytics to Modernize the DoD
The DoD investigates uses for data/workforce analytics to improve training efficiency and enhance force readiness.
With the growing complexity of managing a global military presence on a limited (if large) budget, the US Department of Defense is investing in data analytics use cases to improve efficiency and reduce costs. In addition to integrating previously disconnected databases and developing new systems, this effort will require a change in mentality from a culture that values history and intuition to one that relies on data to produce repeatable outcomes. Long government acquisition cycles of new technology have caused it to fall far behind leading civilian organizations and, while there will be pushback in changing long-standing institutions, significant gains will be possible with relatively little effort.
DoD personnel attend many training courses that can last for up to 1-2 years and cost the government tens of thousands of dollars per attendee in both attendance fees and relocation/housing costs. Candidates are currently screened mentally and physically prior to attendance, as well as reviewed for performance evaluations and time in service. Even with these checks in place, elite courses can have under 40% completion, incurring needless costs and potentially wasting limited spots in the training pipeline that cause active units to go undermanned. With hundreds of available data points spread across DoD personnel files, the first steps are being taken to model likelihood of success. Clear benefits include increased efficiency through a higher overall number of graduates and lower dropout rate that improves force readiness and reduces cost.
Implementation is not without risks. While the current methods of evaluation are dated, subjective (particularly commander’s periodic written evaluations), and inefficient, they represent a status quo with which the military is familiar. Limited training spots at elite schools are coveted and a significant amount of gaming occurs between candidates and their evaluators to fill them. An analytical program will appear to outsiders to be a “black box”, and its outputs will be questioned wherever they differ from existing practices. Furthermore, increased civilian involvement in the demographics of DoD force construction will draw scrutiny should any demographic subset be favored for certain positions. Finally, I wonder to what extent having an “exclusive” pipeline is beneficial at least to unit mentality. Organizations like the special forces pride themselves on being small and extremely selective. It will take a mentality shift to see passing as a “badge of honor” when 80% of trainees graduate instead of 40%.
Beyond training applications, the DoD is attempting to integrate large numbers of currently siloed personnel information databases to improve placement and promotion practices. Officers are currently assigned to fill vacant spots in units according to their time in service and competency evaluations; a system that has been in place since the 1940s. By including hundreds of currently unutilized data points, commanders can fill billets faster with specifically qualified individuals or prioritize training for nearly qualified candidates. Similar to the training application, these changes would increase force readiness and reduce costs.
Given the size of the DoD and its long history, changing the mentality of key leaders will not be easy. These basic applications of workforce/data analytics provide quick, measurable, and significant benefits which would already have been capitalized on in a private-sector setting. Global deployment strains the military’s budget and has a taxing effect on force readiness. Simple changes would allow costs to be reinvested in other programs and candidates to reach their units faster. With the amount of government oversight over the military’s structure and budget, however, any changes will be slow in arriving and receive intense scrutiny.
Sources:
- Lemon, J., 2020. Workforce Analytics Can Provide Benefits Across DOD — Defense Systems. [online] Defense Systems. Available at: <https://defensesystems.com/articles/2018/07/25/workforce-analytics.aspx> [Accessed 13 April 2020].
- Churchill, A., 2020. Data Analytics Will Fuel The Future Of Military Readiness — Defense Systems. [online] Defense Systems. Available at: <https://defensesystems.com/articles/2018/06/19/comment-dod-analytics.aspx> [Accessed 13 April 2020].
- Heckman, J., 2020. Dod Overcoming Culture Challenges To Turn Data ‘Snapshot’ Into Predictive Analytics | Federal News Network. [online] Federal News Network. Available at: <https://federalnewsnetwork.com/big-data/2019/11/dod-overcoming-culture-challenge-to-turn-data-snapshot-into-predictive-analytics/> [Accessed 13 April 2020].
- Nyczepir, D., 2020. How The Army Is Wrangling Its 187 Personnel Data Systems – Fedscoop. [online] FedScoop. Available at: <https://www.fedscoop.com/army-personnel-data-assignment/> [Accessed 13 April 2020].
Thanks, John – you clearly outline many of the reasons why military culture might react poorly to the idea of surrendering decision-making power to an algorithm. One additional reason I might add is that there could be concerns about an algorithm taking away stable government jobs, particularly in towns where the military is the largest employer. How can military leadership best make the case that people analytics can actually improve force readiness?
It was fascinating to learn that the completion rate for elite courses is under 40%… I wonder how that fares with that of other armed forces in other countries? Are there ‘best practices’ (or ‘better practices’) that the DoD can learn from e.g. Israel, which thinking back to the Ment.io case, seems to be quite ahead in the game for using analytics in its military?