The acceleration of information exchange has littered our political, economic, and social landscapes with models once considered robust. Stresses like the global economic crisis and the Arab Spring have forced organizations to rethink decades of management frameworks. These frameworks, from the mundane heuristics we use to choose what to eat for lunch to arcane financial trading programs, are all models. While some laud the big data revolution and data analytics as a panacea, the deluge of information holds as much threat as opportunity. This newfound ability to measure and predict with greater accuracy will increase our reliance on models and allow the Government to attempt projects previously thought impossible. However, it is not the risks that we can measure that will increase the frequency and costs of model destruction, but the risks we cannot quantify. The transformation will not be in the amount of data or our tools to manipulate it, but the increasing velocity with which we must develop and dispose of models to interpret this data. To address this paradigm shift and avoid increasingly frequent crises, the Government must invest in its human capital to prepare its workforce to challenge assumptions, consider counterfactuals, and embrace, not eliminate, risk.
Forecasts are only as good as the people who use them. The power of abundant and cheap information poses ethical, statistical, and philosophical challenges. Will Government workers be equipped to wrestle with the implications of making personal predictions with secret data, or prognostications based on medical diagnostic software? Will regulators be savvy enough to know what a technological crystal ball won’t tell them about future risks? Will a workforce equipped with data-analytics dashboards and predictive models be skeptical enough to investigate underlying assumptions, perceptive enough to distinguish correlation from causation, and agile enough to react when predictions inevitably disappoint? The next transformative issue to impact government will be the ability to grapple with the answers data oracles lay before us.
Government cannot leave it to the engineers and mathematicians to understand the pitfalls of simulations. To wield data with dexterity and confidence the Government must take three steps. First, inspire the Federal space by offering a Federal-wide predictive analytics platform to offer high-profile recognition and lucrative prizes for the best solutions. Second, the Government must make the availability of predictive tools contingent on investing time and effort in training to reframe thinking about analytics. Finally, the Government must enable future leaders by providing them the qualitative skills necessary to assess the underlying weaknesses of models and the communication skills to impart these weaknesses to others. Before Departments are awash in erroneous predictions, sloppy thinking, and the hubris of numbers for numbers’ sake the Government must choose to arm its soldiers with critical reasoning skills. Despite concerns, I am optimistic that America will benefit from the accelerated pace at which we discard and redevelop our tools of prediction. It is only through this process of creative destruction that Government can harness the power of emerging technologies.