Editor’s Note: This article was written in response to our unofficial support for the Department of Defense’s innovation challenge for talent management. We want to help with this effort and have asked for original, creative ideas that reconsider the status quo, shake widely held assumptions, and take on the conventional wisdom about recruitment and retention.
Across the U.S. Army, every day, thousands of soldiers spend hours transferring information out of databases and into screengrabs, and then posting these images on PowerPoint slides. If only one quarter of the Army spends just two hours a week converting data back and forth into analog pictures for one command and staff or commander’s update meeting, that’s over 25 million hours per year wasted. And that’s a conservative estimate. The Army does this, in part, because it lacks many of the systems and the data standards that its own guidance requires. But it is equally because the leader at the head of the table, the commander, may not be comfortable with data.
The strategic documents guiding the U.S. Army are littered with phrases like “data-centric,” “data management,” “data interoperability,” and “data-driven.” The Department of Defense’s Data Strategy states that “Data is a Strategic Asset” and the secretary of the army’s second objective is “…to ensure the Army becomes more data-centric…”. Data is described as a key component to enabling “decision advantage’” and leveraging artificial intelligence and machine learning to out-pace America’s opponents. It is crucial to large-scale joint operations. Data matters to strategies.
The Army is now considering its first millennial recruits, those born after 1981 but before 1997, for brigade commands. At the most senior end, seven of the Army’s current 11 four-star generals are baby boomers. The responsibility for transforming the Army into a data-literate force lies with the commanders bookended by these two generations. In October 2023, the U.S. Army is set to gather over a thousand of its leaders at the annual Battalion and Colonels Command Assessment programs. Here they will select the next generation of the Army’s Command Select List. Candidates will be evaluated on tasks such as how far they can chuck a 10-pound ball backwards over their head and how well they can write a short essay. Noticeably absent from the assessment is any form of data-literacy test.
While the standing power throw has dubious value in determining who should be the next battalion and brigade commander, in the last year artificial intelligence tools like ChatGPT have brought into question essay writing as one of the selection criteria. The Army should take this opportunity to send a clear message about the future: Data literacy is no longer an optional skillset for future officers. The Army should create a mechanism to test candidates for data literacy and offer opportunities for learning at service colleges and in specialized continuing education programs.
Data Matters
The soldiers who the Army selects for battalion and brigade commands have an outsized impact on the culture of their commands. Leaders set the schedules, priorities, and work-life balance of their organizations. If a commander thinks running is important, so will their formation. The way a leader processes information is just as impactful. It sets the standards for how a team uses and organizes its data. As the senior rater of the next generation of leaders, commanders set the tempo of change within their teams. For the last two decades, Army current leaders have slowed the pace of adopting data literacy rather than accelerating it.
The Army’s Data Strategic Effort defines data literacy as “…the ability to derive meaningful information from data so it can be applied effectively to actions and outcomes; it encompasses the ability to read, write, use, and communicate data.” Massachusetts Institute of Technology professor Catherine D’Ignazio and research scientist Rahul Bhargava go further, detailing data literacy as:
- Reading data: involves understanding what data is, and what aspects of the world it represents
- Working with data: involves creating, acquiring, cleaning, and managing it
- Analyzing data: involves filtering, sorting, aggregating, comparing, and performing other such analytic operations on it
- Arguing with data: involves using data to support a larger narrative intended to communicate some message to a particular audience
One can be digitally literate and not data literate, as has been the case with the widespread misunderstanding of crypto currencies. Data literacy is more than just counting things. Instead, data literacy is understanding both how to leverage data for decisions and how to use it to inform analysis and decision-making. It is knowing how to discern good data from bad. Data literacy does not require coding knowledge, though most users trying to parse a data set will find Python immensely helpful. It does not mean an individual needs to know how to build a computer, but instead have a good grasp of Excel, rather than relying so heavily on PowerPoint.
What Is Data?
Data is, at its simplest, structured information. Just because information is digital doesn’t make it data per se. Digital formats just enable speed. A 9-line medical evacuation report offers an example. The 9-line is a schema that translates information into structure. It becomes data. When that 9-line is called by voice over a radio, it must be copied down and then repeated so that the appropriate people can use the data to treat or evacuate a patient. But when it’s data and digital, that’s when a 9-line posted in a chat room can be instantaneously shared in a dozen other chats, all by script, and without transcription errors.
As another example, take the regular situation report. Most units submit their situation reports in raw text, usually in a word document. In my experience, most units’ processes entail multiple echelons copying and pasting text into yet another document, with context and detail falling to the cutting room floor. Even the information that makes it to the higher headquarters often isn’t accessible more than a week later.
An alternative would be to structure these reports as data instead of unstructured text. Give these reports a schema, input them into a common database, and from the moment the data is uploaded, it is organized, tagged, and archived. All the data is available, not just those tidbits each echelon selects. Commanders can query for specific criteria, which adds more data by cataloging how often an event is referenced. The time taken copying and pasting goes back to the soldiers, and a feedback loop of better writing starts working as data input and data output feed off each other.
Traditionally, commanders often turn to staffers or operations research and systems analysts to be the data-literate member of the team. The Army’s Military Intelligence Data Strategy lists “manager, steward, custodian, champion, scientist, architect, engineer, and analyst” as jobs for data-literates. But there is no mention of “commander.” The Department of Defense Data Strategy goes a little further, making it “…the responsibility of all Department of Defense leaders to treat data as a weapon system and manage, secure, and use data for operational effect.” If the Army and Department of Defense want “to fully integrate data across warfighting functions,” then the department should start assessing commanders for data literacy.
The proliferation of “commander’s dashboards” across the services has been an intermediate step in this direction. These dashboards are a data visualization, such as how many soldiers are expired on training. They have helped Army servicemembers see previously missed conclusions and bolster the value in keeping and maintaining good databases. But dashboards are just formulas that answer questions that have already been asked, such as how many soldiers are qualified marksmen. Indeed, once the dashboard can provide the answer, there is little need for the commander to decide anything. Answering new and novel questions from data requires either a commander comfortable with interfacing with data, or a staffer to build them a new dashboard.
When the commander is not data literate, it falls to their subordinates to do double the work. All too often soldiers are stuck transforming data into analog copies (typically on a slide) for presentation to senior leaders. Any enquiry requires soldiers to go back to the data, re-query, and repeat the cycle all over again.
Testing for Data Literacy
There is a simple-to-use test to judge data literacy. The economist Steven Levitt gave up written essays to determine grades for his graduate students years before ChatGPT. Recognizing the need for data literacy, he instead gives students a data set and requires them to clean the set of errors and then draw a novel conclusion from it. The Commander Assessment Program could easily adopt one of the existing data literacy exams, but it could just as easily take a similar approach to Levitt’s test for his students. The Army could provide candidates with an export from almost any Army data system, ask the applicant to clean it up, draw two conclusions from the data, and provide two questions they want more data to answer. This is a small first step, but one which could help the Army to find the more data-fluent commanders needed to change the organization.
Some may argue it is unfair to assess officers for a skill they were never trained to do. Thus, as a first step, data-literacy training should be part of the Army’s professional military education program. The integration of data literacy courses would be a welcome — and overdue — addition to the curriculum at service colleges. However, there is no need for a soldier to wait for their next professional military education course to begin to be trained to be data literate. As the Military Intelligence Data strategy advocates, “…we must become a leader in investing in data literacy training.” The Department of Defense already provides a slew of self-training tools, often partnered with some of the best universities in the nation. Every leader should be familiarizing themselves with at least the fundamentals of data. Self-directed training comes with a time cost, but Army officers will find the skills they pick up will help them save them time several-fold. Soldiers can also start small and simple: Add influencer accounts like Miss Excel to your Instagram feed to get started.
I believe there is a real urgency to start data-literacy assessments because the current generation of professional military education students won’t command for another decade. Thus, it is important to make these changes a priority. Indeed, given the value they place on data, the Army’s Talent Management Task Force should be eager to add data literacy to its assessment criteria. However, this new assessment should, quite obviously, not be the only means to select new commanders. Across the Army, different commands require different amounts of data. Thus, the Army should settle on a standardized baseline and then offer commanders taking command options to expand their knowledge.
Conclusion
The starting horn to transform the Army sounded years ago. While Army current systems stodgily plod along at a gradual pace, artificial intelligence/machine learning algorithms are progressing logarithmically. But these cutting-edge tools need data to run on. Once commands embrace data, it frees up their staff to do more impactful work — and when an entire organization is communicating and moving at the speed of data, the time savings get reinvested in hundreds of game-changing ways. Innovation feeds on time, and it breeds more innovation, and more time savings. The feedback loops are compounding.
On today’s battlefield, leading soldiers requires data literacy. The head of the Combined Arms Center, Lt. Gen. Milford “Beags” Beagle, and his co-authors have examined lessons from Ukraine and argued that the institution needs to evolve. “To optimize our command posts, we must reduce our reliance on the physical dimension (the materiel), increase our utilization of the information dimension (the data), and maximize our relationship with the human dimension (our leaders).” If the Army does not take this seriously, soldiers will die. “To achieve the full potential of convergence, command posts will need to adapt to such an extent that they will be unrecognizable to the generation of leaders that fought in Iraq and Afghanistan.” The Army needs to evolve its systems, and that means adjusting leadership assessment criteria.
Data literacy should hold at least as much weight in selecting commanders as does measuring how far someone can throw a 10-pound ball. Adding a diagnostic assessment during the commander’s assessment program is the perfect wake-up call to both the Army and to the officer corps. When the Army start assessing the future leaders for data literacy, it will finally start to achieve the Department of Defense’s goal of “one Soldier, one byte, and one command at a time.”
Lt. Col. Erik Davis is an active-duty Army officer with over 15 years of experience in special operations. He is also a Gen. Wayne A. Downing Scholar with master’s degrees from King’s College London and the London School of Economics. His assignments have taken him from village stability operations in rural villages in Afghanistan to preparing for high-end conflict in the First Island Chain, and he is contending for command at this fall’s Command Assessment Program. Erik is a 2023 Non-Resident Fellow with the Irregular Warfare Initiative, a joint production of Princeton’s Empirical Studies of Conflict Project and the Modern War Institute at West Point.
Image: U.S. Army photo by Master Sgt. Garrick W. Morgenweck
Comments are closed.