In 2007, the America Competes Act created a new agency called ARPA-E (Advanced Research Projects-Energy) in order to fund and foster breakthrough energy technologies. Since then, one of us, Mark Muro, has consistently endorsed the agency’s vision and strategy. Staffed on a temporary basis by scientists with extraordinary talent for both invention and commercialization, ARPA-E functions like a venture-capital firm with a public-goods focus.

This allows it to fund research and development projects for which the potential benefits to humankind, the U.S. energy system, and commercial gains are large notwithstanding significant technological risk. If a narrow infusion of public funding allows a company to work out the technical and commercial applications of its scientific ideas, the company can pass through the “valley of death,” or at least one of them, and start raising private money and moving toward growth.

Will things actually play out as designed now that ARPA-E is up and running and has made its first grants?

Well, the answer seems to be an emphatic yes, on the evidence of a recent story in New York Times.

In the last 18 months, six of the 37 initial projects ARPA-E funded in late 2009 have attracted private money. To be sure, none of the carbon storage grantees attracted outside investment. But sun and wind power and battery technologies developed in Massachusetts and California and elsewhere did lure investors, with the ratio of private sector investment to public spending averaging roughly four to one. Other companies that may require longer periods of R&D could still be just as successful, if not more so, although with a new bout of fiscal anxiety convulsing the capital ARPA-E’s funding is in jeopardy.

Which is why we thought it was worth taking this opportunity to revisit the rationale for public sector investments in commercial science.

Begin with the fact that it’s well documented that the private sector doesn’t invest an optimal level of money in research and development. Why is this? The basic explanation is that the full value of breakthrough ideas cannot be fully captured by the inventor, and so there is under-investment in invention. Thomas Edison, who has more patents in his name than all but two men in history, went even further. While economic theory suggests that profits from invention are too small to justify adequate expenditure, Edison argued that revolutionary inventions were simply unprofitable.

An 1898 article in Scientific American laid out this view in his words, “The value of a patent to an inventor is directly decreased as the value to the public increases.” The problem, Edison said, is that the large value of breakthrough inventions attracted “infringers,” who reproduce the technology without having to spend the money on research and development. Freed from the burden of having to cover the costs of those initial investments, they are able to sell at a reduced price and undercut the inventor, while the lengthy legal process unfolds.

Yet, according to Edison, for more incremental inventions, where the gains from production are small, the expenses of overcoming trade secrets or paying legal fees on patents are too large relative to the potential gain, and inventors are able to profit. Indeed, according to Neil Baldwin’s biography, Edison spent a great deal of time and money on legal entanglements over intellectual property rights, prompting him to apply for more and more patents on every detailed aspect of his discoveries, which ultimately slowed the spread of their public benefits.

If Edison’s theory is valid, then one can imagine what invention would look like without government backing: plenty of small incremental investments but very few large breakthroughs, and with those that come finding themselves entangled in a web of stultifying legal protections. That is essentially the situation that ARPA-E and similar programs improve upon. By having the public pay some of the bill for the initial research and development, U.S. inventors can compete on a more even playing field with potential infringers from China or elsewhere.

Given that, we suspect Thomas Edison would be delighted with ARPA-E’s initial successes and astounded that the program is facing substantial uncertainty about its future.