It is no secret that, three years since the official end of the recession, many of the world’s economies are still in terrible shape. Here in the United States, unemployment has been above 8 percent for three years. In Europe, public discontent with the state of the economy has caused political shockwaves, most recently in the Netherlands and France. Nonetheless, one of the key problems facing governments today — a problem known as “dynamic inconsistency” — has largely gone unacknowledged outside of the confines of the academic economics community.

The economy depends not only on past and current policy decisions but also on the public’s expectations of future government policies. For example, long-term real interest rates have a far greater effect on the current state of the economy than short-term nominal interest rates do. But current long-term real interest rates are strongly dependent on current expectations of future short-term nominal interest rates — which are determined by the Federal Reserve — and inflation. Thus, the Fed’s ability to influence the economy is primarily due to its ability to shape people’s expectations of its future policies, not from its current policies per se.

These facts result in a problem known as “dynamic inconsistency,” which is most easily illustrated with a non-economic example. Last summer, I worked for a non-governmental organization that has a strict policy of not paying ransoms if one of its volunteers is taken hostage. Imagine that the NGO did not follow this rule, but instead made ransom decisions in a discretionary manner. In this case, the NGO’s optimal strategy would be to pay today’s ransoms (to save the lives of current hostages) and promise never to pay ransoms in the future (to deter future hostage-takers). However, if another hostage situation were to occur, the NGO’s optimal strategy would be, once again, to pay the ransom. Of course, hostage-takers would recognize this inconsistency, so the NGO’s promise to never again pay ransoms is meaningless. However, by committing to a rule to never pay ransoms, the NGO is better able to deter hostage-takers than if it made its decisions in a discretionary manner.

In 1977, economists Finn Kydland and Edward Prescott, who were later awarded the Nobel Memorial Prize in Economics, showed that dynamic inconsistency is pervasive when macroeconomic policies are set in a discretionary manner. For example, if the Fed were given discretionary powers, it would always have an incentive to promise tight monetary policy tomorrow (to keep inflation expectations in check) while adopting a loose policy today (to keep unemployment low). However, the public will realize this inconsistency, and the Fed will face a worse trade-off between unemployment and inflation than if it committed to a policy rule. Even if policymakers are perfectly omniscient and altruistic, discretionary policies will still be sub-optimal.

Over the last four years, the Federal Reserve has continued to face dynamic inconsistency problems, but in reverse. Since we are currently in a “liquidity trap,” the Fed is unable to reduce interest rates to the level needed to keep the economy at full employment. In this situation, the Fed’s optimal discretionary policy is to promise to keep interest rates low even after the economy has exited the liquidity trap, thereby reducing longer-term interest rates today. However, as soon as the economy actually recovered, the Fed would have the incentive to renege on its promise in order to keep inflation in check. Of course, since the public realizes this fact, the Fed’s promise is not credible, and its discretionary policies are not succeeding.

In fact, much of the pre-2008 literature on the liquidity trap was explicitly concerned with designing rules-based monetary policies that would get the economy out of the liquidity trap. These policy rules almost certainly would have produced far better results over the last four years than what we have seen.

While there certainly are reasonable arguments in favor of discretionary policy, it is hard to deny that a decent rules-based monetary policy would have performed as poorly over the last four years as the discretionary policies that were actually enacted.