Despite having a budget for only a few years, the Advanced Research Projects Agency-Energy (ARPA-E) is quickly becoming a microcosm of the larger problems associated with the entire Department of Energy (DOE).

ARPA-E, a program within the DOE, is meant to fund high-risk, high-reward projects that the private sector would likely not fund because the prospects for commercialization are too risky. But in the program’s short existence, ARPA-E is already straying from its mission. As the House of Representatives continues deliberation on the 2013 energy and water spending bill, Representative Paul Broun (R–GA) is offering an amendment that would help return ARPA-E to its intended purpose by prohibiting funding for late-stage research.

Specifically, Broun’s amendment would prohibit any applicant from receiving ARPA-E money if the expected Technology Readiness Level (TRL) is level 7 or higher. Some government agencies use TRLs to determine the stage of technological maturity, which can span from level 1 to level 9. The levels evolve from conceptualization of an idea to a proof-of-concept up through prototype demonstrations and successful mission operations.

The ARPA-E user guide requires that an applicant submit the TRL for the current states as well as the expected state of the technology submitted. It defines TRL-7 as a “system prototype demonstration in an operational environment.”

Broun’s amendment would go a long way in fixing one of the fundamental problems with ARPA-E: The federal government has awarded several ARPA-E grants to companies and projects that are neither high-risk nor something that private industry cannot support. These problems with ARPA-E were recently identified by the Government Accountability Office (GAO), the DOE’s Inspector General, and the House Science, Space, and Technology committee staff.

Of the 44 small and medium-size companies that received an ARPA-E award, the GAO found that 18 had previously received private-sector investment for a similar technology. The GAO found that 12 of those 18 companies planned to use ARPA-E funding to either advance or accelerate prior-funded work.

A large part of the problem is that using taxpayer dollars creates an expectation that the research needs to transition to a market-viable technology. Therefore, it’s easier to pick technologies in later stages of commercialization, and it’s easier for a politician to take credit for a taxpayer-funded technology development and jobs created from opening up a new plant. This is the exact opposite approach needed for ARPA-E as well as the rest of DOE research.

Technologies that lose private financing as they move closer to commercialization are likely the worst bets for taxpayer money, since professional investors have already determined them to be losers. If those technologies are promising, venture capitalists will be eager to invest in them.

Carrying technology from the research and development stage through to commercialization should be a private endeavor. To the extent that the government supports energy research, it should be much earlier in the process.

Congress should hold ARPA-E accountable to its mission, and more scrutiny is necessary to ensure that ARPA-E is not funding projects already receiving private funding or using technicalities to justify those grants.