How to Slow a Charging Hedgehog

According to Philip Tetlock, hedgehogs force data into predetermined models. As one might expect, hedgehogs are frequently wrong, and they learn slowly, in part because they rationalize their poor predictions while leaving their faulty models intact. Hedgehogs tend to be self-assured and overconfident, almost by definition.
      Foxes, on the other hand, adapt their models to fit the data. Foxes are correct more often than hedgehogs, and–not surprisingly–foxes learn more readily. They are more willing to question their models, and they are less likely to succumb to overconfidence.¹
      Despite their biases (or perhaps because of them), hedgehogs are surprisingly influential. They frequent TV news shows and government councils. Leaders often seek the advice of hedgehogs; in fact, leaders may be hedgehogs themselves.
      We could afford to tolerate the hedgehogs’ overconfidence when the world was relatively stable and we enjoyed a forgiving margin of error for our decisions. In the past two weeks, however, the cost of inflexibility has risen considerably. If we fail to adapt and learn now, our resulting missteps could be catastrophic.
      This is doubly true in a world of alert competitors and adversaries. Because a hedgehog tends to adhere to rigid models, a hedgehog is more predictable than a fox. Adversaries and competitors can exploit predictability, and the potential cost of heeding hedgehogs includes increased vulnerability to manipulation and surprise.
      Given these risks, what can we do?
      Clearly, we can’t simply ban hedgehogs from public discourse. The value of this discourse grows with its variety, and now, more than ever, we must entertain a variety of ideas. In any case, the labels hedgehog and fox identify two ideal types or paradigms. Real-world hedgehogs, for example, are not equally stubborn, just as real-world foxes are not equally flexible. Each of us lies somewhere on a scale defined by the two types. Few of us are unalloyed hedgehogs or foxes.²
      That said, we can improve the level of public discourse and decision making, even in a world of overconfident hedgehogs. Doing so is one of the goals of Red Team Journal, and the tools, approaches, and perspectives outlined here are designed to help analysts and decisions makers on this count. When applied thoughtfully and judiciously, these fox-like tools and approaches can debias thinking and facilitate better predictions and decisions, whether the user is a hedgehog or a fox.
      Often, however, formal tools and approaches are inappropriate. When a pressing situation demands action, a leader is unlikely to turn to an analyst and request a Duncker diagram. Nor is an expert commentator likely to draw a morphological box for his or her TV audience.
      At times like these, healthy, assertive skepticism is a worthy substitute for formal tools. Citizens, analysts, and leaders alike should question experts with confidence. Tetlock’s study should strengthen this confidence.
      One well-known questioning technique is the five whys. Each consecutive why helps the questioner and the audience probe an issue more deeply. When using this technique, the questioner can expose hidden assumptions and identify previously unknown links to related causes and issues. Five hows can be just as effective, and a questioner may wish to mix whys and hows in a series of increasingly perceptive questions.
      The technique is simple but powerful. Consider the level of public discourse surrounding the 2003 decision to invade Iraq. Apply five whys or hows and see where the questions lead you. Do the same for the recent financial bailouts and ask yourself whether the discourse in Congress or among the talking heads on TV reached a level of detail and insight consistent with your questions.
      Hedgehogs are quick to answer the call for expert opinion, particularly in times of crisis. As the scale of the crisis grows, so does the penalty of disregarding their overconfidence. We can do our part as citizens, analysts, decision makers, and leaders by exercising assertive skepticism. Few things can slow a charging hedgehog like a string of five well-timed whys.

Further Reading:

Notes:

1. See Tetlock’s 2006 book Expert Political Judgment: How Good Is It? How Can We Know?. [↩]
2. Even so, it is possible to calibrate a decision maker’s degree of overconfidence. [↩]