The National Science Foundation (NSF) has just released preliminary results of its second annual Business R&D and Innovation Survey, which they call the BRDIS (it has to be one of the least catchy acronyms I’ve ever encountered; the NSF has a history of creating uncatchy acronyms, but this one, unfortunately, makes me think of the word “bris”). Here’s a quick summary from their web page:
Preliminary figures from the Business R&D and Innovation Survey (BRDIS)–fielded for the first time in 2009–indicate that some 22 percent of companies in manufacturing industries reported one or more product innovations (goods and/or services) in the period 2006-08 and about 22 percent introduced process innovations (new methods for manufacturing or production; logistics, delivery, distribution; support activities). Among manufacturing sectors with the highest incidence of innovation (computer/electronic products, chemicals, and electrical equipment/appliances/components), 37 percent to 45 percent of companies reported product innovations and 28 percent to 34 percent reported process innovations. About 8 percent of the estimated 1.5 million for-profit companies represented by the survey are classified as manufacturing industries.
The vast majority (92 percent) of companies represented by the survey are classified as nonmanufacturing. Here, 8 percent of companies reported product innovations and 8 percent reported process innovations. Largely, these companies are in such industries as wholesale/retail trade, hotels, entertainment, and personal services, where the rate of product and process innovation is low. The information sector was an innovation standout among nonmanufacturing industries, with 30 percent of companies reporting product innovations and 20 percent reporting process innovations.
Companies that perform and/or fund R&D reported a far higher incidence of innovation than did companies without any R&D activity: 66 percent of companies with R&D activity reported product innovations and 51 percent reported process innovations. In contrast, 7 percent of companies that were not active in R&D reported product innovations and 8 percent reported process innovations.
When all of these industry groupings are taken together, about 9 percent of companies represented by the survey were product innovators in the period 2006-08 and about 9 percent were process innovators. (Product and process innovation overlap to some extent, and the incidence figures are in general not additive.)
The National Science Foundation (NSF) has just released its first ever Business R&D and Innovation Survey (BRDIS). Developed jointly with the U.S. Census Bureau, it reveals that companies located in the U.S. spent $330 billion on R&D in 2008, with $234 billion of that used for research in facilities located in the U.S.
NSF’s previous R&D instrument, carried out every year since 1953, was called the Survey of Industrial Research and Development. But research is conducted differently today than it was 50 years ago, and back in 2004 the National Academies’ Committee on National Statistics recommended that NSF develop an updated version. New features surveyed include:
- worldwide R&D expenses
- R&D employee headcount by occupation category
- R&D expenses by detailed business segments
- share of R&D devoted to new business areas and new science or technology activities
Companies with R&D reported a high ratio of domestic U.S. sales to worldwide sales: well over 60 percent.
The data were gathered from a representative sample of 40,000 U.S. businesses (including U.S. owned and also U.S. affiliates of companies owned outside the U.S.).
I’ve just read a July 2008 report of a National Science Foundation conference, that brought together a panel of experts to try to figure out a way to measure innovation (click here for the full report). They started with the observation that we don’t know very much about how to measure the outcomes of innovation, and we don’t know very much about what characteristics make the difference between a successful or failed innovation.
Briefly, they identified four areas that businesses and governments could measure.
- Measure what innovation is. Determine the rates of return over a project’s lifecycle, and how this looks different for different kinds of innovation.
- Study how and why innovation takes place, by examining the organizational structures and cultures of successful innovations (this is the focus of my own research).
- Figure out how to capture the complete range of consequences of innovation, including the downsides of “creative destruction”– unemployment, regional winners and losers, and social responsibility.
- We need a better understanding of the complete environment of innovation, from globalization to government IT law to tax policy to technological change, all is interconnected.
And then the panel had four recommendations:
- Pursue new directions in data gathering measuring the net outcomes of innovation. This includes data on innovation processes within organizations.
- Pursue new directions in measuring the net inputs (costs) associated with innovation.
- Gather data on innovation processes that span across organizations. (I like this one, because it acknowledges my claim about “collaborative webs”: that innovation is rarely bounded within the walls of one company.)
- New information gathering and management techniques could allow us to confidentially aggregate data from many companies, and share it with everyone.
I looked at the list of participants and they all seem very qualified–with expertise focusing on economics, consulting, workforce and human resources, computer science, and business strategy. I didn’t see any of the top creativity or innovation experts, however, which was a bit puzzling. Innovation experts absolutely must be deeply involved in all of the steps recommended by this report, I’m not sure why they weren’t involved at this early stage.
Also see my post of March 4, 2008 about the www.innovationmetrics.gov panel of experts.
I just returned from speaking at a workshop hosted by the National Science Foundation headquarters in Arlington, Virginia. I know everyone loves to bash government bureaucracy, but the NSF is a quality organization and I’m always impressed with everything they do. This workshop was no different, with the title “Art, Creativity, and Learning”. The mission our group of experts faced: to prepare a list of important research questions for the future, and to advise the NSF on what types of research should be funded in the next few years.
The event was organized by Christopher Tyler of the Smith-Kettlewell Institute in San Francisco, and neuroscience research was a constant theme–what does the brain look like when it’s being creative? Or when it’s listening to music? Or looking at a painting? The other constant theme was, when we participate in the arts or in creative pursuits, do we learn things that can make us smarter in general? For example, everyone seems to believe that playing music makes you better at math. But, surprisingly, there’s no solid evidence that’s true. We proposed several research projects that could help us to understand what’s uniquely valuable about the arts.
For me, the high points of the conference were presentations by Ellen WInner of Boston College, perhaps the leading scholar asking questions about the arts, development, and learning; and Dan Levitin, a neuroscientist and author of the best seller This is Your Brain on Music.
I wish I could report some surprising new answers, but our goal was to ask the big unanswered questions, and we did a good job of that: Does participating in the arts give you any increase in general mental ability that transfers to others domains? If you use dance, music, or painting in math or science class, does it help people learn math or science better? (This is a common belief that has no solid research support.) I personally love the arts and I want them to remain in the curriculum. But, as a scientist, I want to be able to argue for the arts using solid data and research findings, not just wishful thinking.