Earlier this summer, Washington Monthly ran a piece by Haley Sweetland Edwards entitled “A Short History of Data-Driven Government.” The article focuses on a decidedly dispiriting story about Robert McNamara, who she calls the “father of data-based government.”
Edwards first describes how McNamara, a former army captain, learned about efficiency at Harvard Business School, used his skills to climb the ladder at Ford Motor Company, and eventually landed in President Kennedy’s cabinet, as Secretary of Defense.
Here’s where the tale turns cautionary. Edwards writes that “McNamara applying those same ideas about efficiency, instituted the now-infamous ‘body count policy’ in the quickly escalating Vietnam War.” This policy, she explains, was later “roundly maligned by liberals and conservatives alike, underscoring the problem of data-driven policies that, by failing to quantify holistic goals, sometimes incentivize appalling behaviors.”
Edwards describes how data-driven governance soon faded out of fashion, until President Clinton revived it in the 1990s, training federal agencies how to measure success and create progress reports. Many of these ideas have carried through into the Obama administration, which requires a similar type of reporting from federal agencies. These efforts have seen success. Edwards mentions that, under Clinton’s initiative, the Coast Guard, “using data to map the problem of crew casualties in the towboat industry, was able to reduce fatalities by two-thirds from 1997 to 2000.” And under Obama, the Treasury’s Bureau of Engraving and Printing converted “to a centralized accounting system—a move that the Financial Management Service estimated could save the government $400 million to $500 million annually.”
Edwards cautions, however, that no “data-driven performance management tool” is “immune to the pitfalls made infamous by McNamara” because metrics are so easy to game.
I wish Edwards had done more to derive lessons from McNamara’s downfalls. The problem was a defective end-goal, not the data analysis itself. Thus, perhaps the biggest takeaway from revisiting this story is the reminder that data analysis in government cannot be an end in itself, the goals must be right-minded. This, of course, is always the biggest challenge in government—deciding which goals are right-minded. In business, the objective is typically easy to define: create sustainable profitability. In government, the objectives are much more complex to formulate, especially as one scales up from local to national governance, where even reaching acceptable goals is often an impossible feat. But as McNamara’s story shows, even with the “big data” available today, this process of defining goals must remain demanding because what we’re working toward makes all the difference.