The US government gets a very high return on government investment in scientific research at universities and research institutions, according to calculations by the MIT Technology Licensing Office (TLO). While research institutions and inventors share royalties worth slightly less than 3 percent of sales of licensed technologies, the government gets back about 15 percent of sales through income taxes, payroll taxes, capital gains taxes and corporate income taxes. In addition, the government has free use of any patents.
The government payback is estimated at six times the royalties that businesses pay to universities and research institutions, based on a very conservative estimate that does not include secondary economic effects such as the business generated when new employees make purchases from their salaries. The research institutions receiving royalties are required by federal law to share the royalties with the individual inventors, and the university's share must, by law, be plowed back into its research and educational activities.
The TLO analysis "illustrates a fundamental misunderstanding of the Bayh-Dole Act of 1980 seen in recent news reports claiming that the government doesn't get anything back from government-funded research," said Lita Nelsen, director of the TLO since 1986.
"The Bayh-Dole Act is working precisely as intended, developing the fruits of government-funded research for the benefit of society," she said. "Economic development through exploitation of intellectual property is now widely considered one of the major benefits of federally sponsored research.
"Only government can make, for dozens of years, the patient investment in basic science needed for scientific discoveries. Only then can private business afford to take the considerable risk of licensing the patented discoveries and investing millions of dollars more to develop the technology into a device or medicine that will bring great benefits to society. This process generates new jobs, new companies, sometimes new industries and new wealth. Government then gets back its share through taxing that wealth," while the public benefits not only from the returned taxes, but from the new products, new companies and new jobs that result, Ms. Nelsen said in an interview Monday.
The Senate Judiciary Committee, in discussing the Bayh-Dole bill in 1980, decided that the government should not participate in the direct income from licensing but instead get its return from taxation on the resulting increase in economic activity.
TAX PAYBACK
The TLO analysis is based on a national study by the Association of University Technology Managers (AUTM) of technology licensing in 1996 through 173 major universities, hospitals and other research institutions. AUTM found that the research institutions received $514 million in royalties on $20.6 billion of sales of licensed products and that the total economic activity from licensing supported 212,500 jobs.
The TLO analysis made a conservative estimate that the tax payback from the sales totaled $3 billion, including $1.2 billion in federal income taxes; $1.5 billion in Social Security and Medicare contributions from the employer and employee; over $100 million in capital gains taxes from investors in stocks of start-up firms; and over $200 million in corporate income taxes, assuming a very conservative 3 percent profit as many of these companies are not yet profitable.
PATENTS: FROM 300 TO 2,200
Ms. Nelsen, commenting on the success of the Bayh-Dole legislation, noted that the number of US patents granted to American universities in a year rose from about 300 in 1980 to 2,200 in 1996. More than 1,900 companies have been formed through licensing activity since the inception of the Bayh-Dole Act in 1980.
Universities and research institutions, she said, have developed policies to control the potentially conflicting interests of universities -- dedicated to dissemination of knowledge -- and companies, which often need to keep information proprietary for competive reasons. MIT is acknowledged to have one of the strictest policies on managing these and other conflicts of interest arising from its licenses and collaborations with industry.
Research institutions have been much more successful than the government was in licensing patents. Before Bayh-Dole, the government by 1978 had licensed only 4 percent of the 28,000 patents it owned, whereas successful universities now license nearly half their patents. Prior to Bayh-Dole, there was no market force to encourage licensing, and research was frequently not developed into new products.
"A key aspect of licensing of inventions under Bayh-Dole was the ability to grant an exclusive license," she said. "How could the federal government justify allowing a single company to be given the advantage of intellectual property developed under taxpayer funding? Because exclusive licenses were imperative for the development of early-stage technology. The commercial licensee must devote substantial time and money to attempt to develop the technology, with no guarantee that it will be successful.
"Exclusive licenses are an inducement and reward for a company willing to step forward and take such a risk -- knowing that if it succeeds in the development, the exclusive license will protect it from having its product copied by those who weren't willing to take that risk."
Institutions sometimes share that risk by taking shares of stock in start-up companies as a form of royalties. "This practice was highly controversial initially, but it has become accepted as experience is gained and the predicted disasters have been largely averted through thoughtful formation and enforcement of policies," Ms. Nelsen said.
For universities, she said, "the direct economic impact of technology licensing has been relatively small -- a surprise to many who believe that royalties could compensate for declining federal support of research. Because of the high costs of patenting, most university licensing offices barely break even.
"But there are other benefits. The local results of university licensing -- new local businesses, jobs and real estate and sales tax revenues -- have sweetened town/gown relationships a bit.
"Another unpredicted effect has been to motivate students to become entrepreneurs and to increase their awareness of the potential commercial and social utility of their research findings. Many engineering, design and business development courses now include a class on patenting and technology transfer, and entrepreneurship courses are very popular in business schools," Ms. Nelsen said.
A version of this article appeared in MIT Tech Talk on April 15, 1998.
