Over the past few years, Congress has taken aim at U.S. science funding. The present America Competes Reauthorization Act is but one example. This bill would make several changes to science funding, but two stand out. One, it states that "the results of any research, development, demonstration, or commercial application projects or activities of the [Energy] Department may not be used for regulatory assessments or determinations by [f]ederal regulatory authorities." Two, it substitutes politicians' preferences for scientists' when deciding what research gets funded. Neither change is a good idea, and the bill as a whole is detrimental to American competitiveness.

To see why, consider the analogy of home repair. Let's say your old electric hot-water heater breaks and you call in a plumber to install a new one. Typically, the plumber would ask you some questions and then provide options. You'd choose your preferred one, with your choice limited by your budget. You'd probably also try to pick an experienced plumber so as to minimize the chance that you'd get bad advice. While you'd try to hold the line against cost overruns, you'd likely let the plumber decide how to spend your budget.

Now consider the same process under the logic of the America Competes bill. You'd call a plumber, but you'd cut off any discussion about options. After all, these would be "assessments" based on years of prior research and development, and this is not allowed. Instead, you'd choose the type of heater that had the best advertisement, or the one made in the plant where your friend works. Then the plumber would try to talk about an overall budget, but you'd cut off that discussion, too. Instead, you'd hand the plumber an itemized list of what you believe each component of the heater should cost. You'd handle potential cost overruns by not letting the plumber deviate from your list.

Now, maybe you are a plumber yourself and the second process seems right to you. That's great. But most of us aren't plumbers, and have limited ability to, say, install a new gas water heater. Most of us would be quite happy to have someone knowledgeable describe the options, and most of us would not micromanage the plumber's choice of pipe fittings based on some prior belief that Teflon tape or copper ferules are somehow not part of "real" plumbing.

You could substitute doctors or electricians or lawyers or any one of a thousand other professions for plumber in this analogy. The basic point is the same: In our daily lives, when we need something done that we can't do ourselves, we seek out the best people we can afford, listen to their advice and give them enough leeway to use their expertise to our advantage.

This is not just common sense. It is also basic economics. We delegate tasks that we want done but can't do ourselves, such as replacing a water heater or developing new science. Two major problems exist in delegation: picking the right people and limiting shirking. To pick the right people, we rely on signals such as experience, track record, education and so on. To limit shirking, we provide incentives that help induce the people we hire to want what we want. For example, we might pay based on performance, or provide equity in a company. The logic is straightforward: If we can get those we hire to want what we want, then they will act in our best interests without our having to compel them to do so.

The present system of science funding is good at solving both of these delegation problems. Thanks to records of publications, grants and training, granting agencies such as the National Science Foundation (NSF) have access to lots of information about those who would carry out the science. NSF also has the expertise to assess all these signals and so increase the chance of picking the right people. And the scientists who get grants have pretty good incentives to act in society's best interests as well. They benefit most when their grants are successful, as this provides career benefits and increased chances of getting more grants. If society values science, funding the present system as it is provides ample incentives to produce science. It also allows the flexibility for scientists to pursue the most promising new areas of research, which aren't always apparent ahead of time.

America Competes reduces these incentives and reduces expert ability to gauge the quality of research and identify new areas of research. While Congress has both the right and the responsibility to manage science funding, it is far better to continue to fund government agencies such as the NSF as a whole, adjusting their overall budgets as society's relative need for scientific progress changes. In this way, Congress can continue to serve its valuable role without destroying the incentives that have made the United States a leader in scientific innovation — innovation that has led to vast improvements in both our understanding of the world and in our quality of life.

Siegel is an associate professor of political science at Duke University.