Given the intense competition for faculty jobs and federal research funding and the resulting need for most Ph.D. holders to seek careers outside academe, it was good news when our colleague Jeffrey Mervis recently reported a “significant rise in corporate funding of fundamental science since 2012.” Just weeks later, however, experts in the economics of science gathered in Washington, D.C., for a conference entitled “The Decline in Corporate Research: Should We Worry?”
Both narratives, it turns out, are true. At the conference Ashish Arora, a business professor at Duke University in Durham, North Carolina, laid out strong evidence for a “fairly broad-based cross-sector decline” from the peak of corporate research spending in the last century. Whether the much more recent rise in R&D investment that Mervis reported is an “uptick or the beginning of a trend” is not yet clear, noted conference participant Jeffrey Smith, a Ph.D. biophysicist with an MBA who heads the health care practice of McKinsey & Company. This much is certain, however: A decades-long shift has transformed corporate research from a landscape dominated by large, stable, well-funded scientific institutions focused on long-term employment into today’s much more fractured and fluid innovation ecosystem, with significant career implications.
Basic and applied science
The research establishments that U.S. industrial giants formerly maintained—such as the DuPont Experimental Station and the American Telephone & Telegraph Company’s Bell Laboratories, to name just two—created such epoch-making inventions as synthetic fibers and the transistor. They also did groundbreaking basic science that garnered multiple Nobel Prizes, science historian David Hounshell of Carnegie Mellon University in Pittsburgh, Pennsylvania, recalled at the conference. Starting with the founding of the nation’s first great corporate research center, General Electric Research Laboratory in Schenectady, New York, in 1900, and followed by major labs at DuPont, AT&T, Kodak, and numerous other leading companies, these institutions were inspired by the prestigious labs of major 19th century German industrial firms. (At the time, German science led the world and “everyone had a German Ph.D.,” Hounshell said.) The American institutions upheld the “pure science ideal” imbued by German universities, so, in addition to working on new products, they pursued basic research and publication in prestigious scientific journals. For decades, General Electric’s slogan proclaimed that “Progress is our most important product.”
But, conference participants noted, the history of the great corporate labs also reveals how the relationships among pure and applied science and research and development are far more complex than those rubrics suggest—a point that is still relevant to scientists today. In reality, ideas and knowledge do not lead in a single direction from fundamental science through applications that result in commercial products. Instead, questions and problems arising out of practical issues can circle back to fundamental discoveries, which can then lead to additional applications and discoveries.
For example, recalled mathematician Ralph Gomory, now at New York University in New York City but formerly IBM’s senior vice president for science and technology, two IBM scientists, Gerd Binnig and Heinrich Rohrer, shared the 1986 Nobel Prize in Physics for inventing the scanning tunneling microscope, which they did on company time as part of an effort to understand silicone. In another case, Bell Labs radio astronomers Arno Penzias and Robert Wilson stumbled onto a groundbreaking discovery about the origin of the universe while trying to get rid of “static” that was spoiling their radio antenna observations. After eliminating bird droppings and ruling out military and commercial transmissions and every other potential cause they could think of, they realized that the “noise” had to be the predicted—but never before observed—cosmic microwave background (CMB) radiation that is a remnant of the big bang. Penzias and Wilson’s antenna difficulties in the New Jersey countryside thus led them to the 1978 Nobel Prize, also in physics.
The allure of basic research also served as a powerful tool for recruiting the talent companies needed to advance their technologies, executives believed. For example, a North American Aviation document about the founding of the company’s science center, as quoted by Hounshell, stated that “To attract … top-notch research men, the company must display an enthusiasm for good science as an end in itself. It is generally true that the degree of company-oriented enthusiasm shown by the scientist is directly proportional to the degree of science-oriented enthusiasm displayed by the company. … The enthusiasm must be felt; it cannot be produced by edict.”
Today, many of those fabled institutions are gone or greatly diminished, thanks to a complex combination of trends over a period of decades. The rapid and enormous growth of federal funding for university research during the late 1940s and 1950s—the heyday of the great corporate labs—soon raised U.S. academic science to world preeminence. (As Arora said, “Germans now come here to get Ph.D.s.”) Ultimately, it also helped reshape the U.S. innovation system. Another contributor was a change in patent law regarding federally funded research that Congress enacted in the 1980s. In addition, at around the same time, executive compensation shifted away from salaries and company pensions—which encouraged managing firms for the long term—toward stock options, which place more emphasis on shorter-term stock value, said economist and entrepreneur William Raduchel of Georgetown University in Washington, D.C. This encouraged interest among researchers, both academic and industrial, in monetizing their own work rather than remaining employees of large corporations that reaped the benefits. Some scientists began to forge careers spanning entrepreneurship and academe in ways that were formerly either not possible or uncommon; such options and opportunities are likely to continue expanding.
These days, the basic research that fuels breakthrough science and supports important technological innovation mostly happens at universities. Products that once would have emerged from basic R&D by company scientists and engineers now frequently come to market through innovation chains that include university research followed by initial development at startup companies supported by venture capital and further development by established firms that absorb either startups’ ideas or the startups themselves.
On an individual level, scientists who once might have spent their working lives with a single large, long-term employer that provided ample salaries, benefit packages, and pensions now often take on relatively short job commitments during careers consisting of multiple different jobs and employers or become entrepreneurs. Importantly, conference participants noted, this means that economic and employment conditions and structures, including those that shape scientific careers, now vary considerably among industries, scientific fields, company types, and even regions of the country. Understanding the differences and the specifics of various arrangements is crucial to finding career opportunities. Campus-based discussions, however, often fail to capture the variety and complexity of corporate scientific work available today.
Just as conference presenters reached no conclusions that apply to all of corporate science, scientists hoping to succeed in that world should take nothing for granted. Instead, in today’s highly entrepreneurial environment, they need to arm themselves with as much specific and current information as possible as they look for work in the corporate world. They can, among other things, talk with people who work in the field they’re interested in, attend career events and scientific or trade meetings focused on that field, read trade publications that cover the relevant industry, study the websites and documents of specific companies, and—if possible—do an internship or other participatory project at a company. Simply put, they need to deploy the same ability to ask searching questions and seek out accurate answers that they have spent years developing in their university labs.