Market-Based Executive Power
The possibility that undemocratic regimes might use prediction markets to improve a head of state’s ability to monitor decision making shows that prediction markets can assist in executive as well as in legislative and judicial decision making. But could prediction markets in principle supplant executive decision making? Surely they could perform some tasks commonly viewed as executive, for example, with predictive cost-benefit analysis or a text-authoring market used to enact regulations. Other tasks, such as the power to veto legislation, are not present in all governmental systems anyway. And still others might be outsourced with the help of prediction market mechanisms. For example, one might use prediction markets to decide which civil and criminal cases to bring, and one might use market-based reverse auctions to assign these tasks to particular agents, whether individuals or private groups (see Chapter 5). Whether or not such changes would be desirable, it might still seem that there are many decisions that must be left to the power of a single chief executive. One reason might be that there are accountability benefits. This might justify great executive power in our present scheme of government, given the danger that special interests might influence decision making. Special interests can influence the president too, but it may be easier for the public to monitor that individual’s good faith. Prediction markets, however, might minimize the effects of special interests, making this justification somewhat less powerful. Presumably, we care about accountability not because we want to punish bad decisions but because we hope that the prospect of such punishment will lead to good decisions. Whether the incentives that accountability provides are greater or less in magnitude than the financial incentives provided by prediction markets is an empirical question. Another reason is that the government must conduct some decision making in secret. Prediction markets, however, might be used to delimit who has access to different classes of secret information. Those individuals might make recommendations based on their access to secret information without revealing it. Prediction markets in turn might make decisions based on these individuals’ reputations and based on their statements of how strongly they feel. While government currently delegates power to individuals who have access to confidential information, the prediction market system would provide flexibility for participants to discount those individuals’ assertions. Alternatively, one could imagine allowing only individuals with access to classified information to trade on (and later, to resolve the payouts in) a prediction market. This approach gives a direct financial incentive to the individuals with access to that information. The strategy is analogous to the proposal considered in Chapter 2 of allowing individuals with access to classified information to trade on prediction markets so that the government can make credible predictions to the public. A related concern, applicable particularly in the military context, is the need for surprise. If the object sometimes is to surprise an enemy, it does not make sense to debate and announce that surprise on the Internet. For this problem, the first of the strategies discussed above will not work, but the second would. The ex post decision maker would assess the wisdom of any surprise decision, but the prediction market would approve of a surprise decision only when it seemed wise to have a surprise action. If it makes sense sometimes to make decision making truly random (as it probably does for a baseball manager or football coach who aims to keep the opposition off guard), then the prediction market might predicate a decision on consultation of a random number generator. The market in turn would anticipate whether such consultation was wise, not whether it was good that a particular number was generated. All of these concerns are relevant, of course, not only in imagining the unlikely prospect of market-based government but also because many private organizations rely to some degree on executive power. Many may feel that an organization needs a leader, though this may sometimes be a reflection of weaknesses in alternative existing decision making approaches. Accountability, secrecy, and surprise are all reasons why public prediction markets that vet all decisions might not be in an organization’s interests. But even in these cases, prediction markets can limit the degree to which a single individual is invested with power. And if it is wise to assign some decisions to individuals, the same person need not always be assigned the same decision. Prediction markets can determine whether it is and who should be granted which parts of executive authority.
Leave a Reply