Concerning Federally Sponsored Inducement Prizes in Engineering and Science (1999)

important yet seemingly intractable challenges might serve to legitimize promising new technological approaches, increase a researcher's or contestant's chances of long-term funding, or serve as an important “signal” to venture capitalists or other sources of private funding. Moreover, prize contests of this type may also serve to attract public attention to neglected societal challenges and generate public support for additional research and policy experimentation related to these challenges.

• Build “social capital.”²² Contests can stimulate the capacity of individuals and groups to work together for mutual benefit. Social capital is strengthened through the collaborative aspect of incentive programs—the activity of learning inspired among those who form teams or interdisciplinary groups to compete. A contest, unlike a procurement contract, is likely to lead to the formation of new, ad hoc partnerships whose members determine leadership and direction with specific goals in mind.

For example, the Royal Aeronautical Society's Kremer Prizes, offered two decades ago, attracted a group of engineers at MIT to form a team and design an entry. Their entry was successful in the latter stages of the competition, and the same team went on to conduct the Daedalus Project, whose human-powered aircraft established virtually all current world range and endurance records, notably a flight of 72+ miles between the Greek islands of Crete and Santorini in 1988. The core of the Daedalus team has evolved into a commercial enterprise, Aurora Space Sciences, whose current mission is to develop affordable robotic aircraft, primarily for high-altitude atmospheric research. Thus the Kremer Prizes focused and advanced the careers of participants in unexpected directions.²³

In summary, the history of inducement prize contests demonstrates that such contests can serve a broad range of objectives—some highly specified, others very broadly defined. Regardless of their stated primary objectives, many inducement prize contests in science and technology place great emphasis on public education and inspiration as a major goal. Moreover, as the discussion of potential prize objectives makes clear, prize contests can be designed to stimulate effort across the entire spectrum of research and innovation activities, including basic research (Wolfskehl Prize in mathematics),²⁴ technology development (the longitude prize), technology diffusion (the SERP prize), as well as managerial/organizational innovation (the Baldrige Awards), etc.

Inducement prize contests usually fall into one of two basic categories: best-entry contests, which reward the best solution within a given time period, and defined-objective contests, which may remain open until a specific goal is reached. One example of a best-entry inducement prize contest is the privately funded Loebner Prize, which each year gives a cash award and a medal for the computer that gives the most “human” responses to questions.²⁵ Another best-entry prize might reward the development of toys that stimulate scientific learning in children, an important educational goal of the nation. Examples of defined-objective contests are aviation prizes such as the aforementioned Oertig and Kremer Prizes.

Recent space prizes such as the X Prize and the CATS Prize also fit this model. Compared with goal-oriented prize contests, best-entry prize contests are likely to require a more complex and subjective judging process to choose the winner.

Case studies of specific prizes presented at the workshop, as well as a growing body of research on contests, grants, procurement contracts, patent races, and the optimal design of federal R&D programs,²⁶ suggest that the following guidelines may prove helpful in structuring specific best-entry or goal-oriented inducement prize contests:

• Contest rules should be seen as transparent, simple, fair, and unbiased. Goal selection must be transparent and credible, the criteria for winning must be clear, and the process for determining winners must be perceived to be fair and unbiased. Clearly this represents much more of a challenge to prize contests targeted at large, complex, societal challenges, than to those that are focused on more readily quantifiable or definable technical objectives.

• Prizes should be commensurate with the effort required and goals sought. For example, a prize contest for the design of the best educational toy might offer a modest prize, given the relatively low investment needed to enter. On the other hand, rewards (financial and other) for prize contests with more ambitious objectives—such as the development and marketing of super-efficient refrigerators —must be significantly larger in order to attract qualified contestants.

At the extremes, if the value of a prize is too small relative to the cost of competing for it, it will attract no contestants. On the other hand, if a prize is much larger than the anticipated cost of competing for it, the contest could draw too many contestants. This would lower the probability of winning the prize for any given entrant, and reduce the expected payoff. This would also raise the cost of administering the prize, i.e., the cost of reviewing and filtering large numbers of prize entries. While there may be ways to reduce the costs associated with singling out the highest performing contestants (e.g., via contestant auctions, entry fees, and other mechanisms),²⁷ excessively large prizes may affect contestant behavior in ways that reduce the effectiveness of these mechanisms. Furthermore, it might lead to excessive duplication of effort. Indeed, sponsoring a prize that is much larger than the expected cost of competing for it makes sense only if the sponsor believes that there are a large number of very different technical approaches that might work, and so wants to get a large number of contestants participating in the prize competition.

The closer the objectives of an inducement prize contest lie to perceived market opportunities and the existing capabilities of would-be contestants, the lower the costs of competing for it will be, and the smaller the prize needs to be to attract competitors. Conversely, the further a contest's objectives lie from perceived market opportunities (high-risk challenges far beyond the current technological horizon, or otherwise neglected technologies or societal challenges), the higher the intended inducement effect will be, the higher the cost of competing for it will be, and the larger the prize must be to attract contestants.

• Treatment of intellectual property resulting from prize contests should be properly aligned with the objectives and incentive structure of the prize contest. This issue of awarding intellectual property rights (IPR) must be considered carefully in designing prize contests. No one model or approach will fit all contests. In some cases, contests that invite firms to develop new technologies might be expected to leave the rights with the inventor. In others, intellectual property ownership might be tilted in different directions according to the size of prizes and the intent of contests. In certain cases, the property rights associated with a prize-winning entry might be placed in the public domain, in which case the cash or other non-IPR-related rewards would need to be much larger. In short, the best IPR policy is one that matches the objectives and incentive structures of particular prize contests.

Ultimately, the administering agency or other sponsor should determine the goal of each contest in light of its mission objectives, the overall objective of the research area involved, and the magnitude of the R&D challenge required to win the contest.

While this report is aimed primarily at federal agencies, the same principles of prize contest design and administration can apply to inducement prizes funded or administered by the private sector. In terms of administration, it is logical to expect a range of models for contests, including:

• Agency funding and administration

• Private funding and administration

• Joint agency-private funding and administration

• Private funding, agency administration

For federal agencies to fund inducement prize contests, Congress (congressional committees) would have to develop a mechanism to authorize and appropriate money that might not be spent for several years. At a time of great need, however, unspent federal funds could be difficult for prize-sponsoring agencies to retain. Obviously, even the best-designed prize contests will be futile unless agencies can guarantee access to prize money when the winner steps forward.

Some agencies—depending on the importance of research to their mission objectives —may be able to guarantee prizes autonomously, especially when prize amounts represent a small percentage of the research budget. A more general solution might be an endowment mechanism by which federal prize money could be reserved until claimed.

Prize contests funded by nonfederal sources would not be subject to this uncertainty. The Department of Commerce's Malcolm Baldrige National Quality Awards program is an example of a privately funded, agency-administered model, where a privately created foundation offers stable, long-term support. The use of private-sector judges brings credibility and reduces political influence on the selection process. At the same time, government participation adds prestige and a sense of fairness.