On Wednesday, we gathered a group of experts to discuss Donald Trump’s priorities in science. Now it’s Hillary Clinton’s turn.
After studying C-SPAN transcript data, we came up with three topics that represent Clinton’s science priorities: education, advanced manufacturing and funding scientific research as a way to grow the economy and create jobs. (Trump’s were military technology, energy and space.) The subjects also show up in the answers she gave to Science Debate, an organization that asks candidates about a range of science-related policy issues. We’ve brought together a roundtable of experts in those fields who will talk about the policy questions they see as most important, how their fields interact, and the ways that science and politics influence each other
Our participants are: Erica Fuchs, professor of engineering and public policy at Carnegie Mellon University; Elizabeth Mann, a fellow at the Brown Center on Education Policy at the Brookings Institute; and Maryann Feldman, distinguished professor of public policy at the University of North Carolina and the director of the National Science Foundation’s Science of Science and Innovation Policy program. The moderator is Maggie Koerth-Baker, senior science writer at FiveThirtyEight.
(The chat below has been lightly edited.)
Maggie: I’m going to start this off with the same question that I started the Trump chat with: What connections do you see between your fields? Why might a politician who is interested in education also be interested in advanced manufacturing and also be interested in using basic science research as a way to jump-start economic growth and job opportunities?
Erica: At least in the work that I do (advanced manufacturing), these three are all intimately connected — advanced manufacturing is by definition new and at the cutting edge of science. Often we are slowed in bringing new materials and processes to commercial reality due to our lack of scientific knowledge. Funding of research can be critical to overcoming the notorious “valley of death.”
Maggie: What is the valley of death?
Erica: The valley of death is the [roughly] 20-30 years it takes from an initial scientific discovery until that discovery makes it to market. Especially in manufacturing industries and in “physical products” (not software) this valley of death can be difficult to overcome due to the large amount of funding required to commercialize manufactured technologies. Venture capitalists today can find these large investments too risky.
Elizabeth: I think that with technological advances, we’re going to see big changes in the structure of the labor market in the near future. This, in turn, has important implications for the types of preparation, skills and education that politicians need to think about providing for students early on.
And to the point of creating a strong education system and educated workers, many policymakers are focused on making sure that decisions about how to provide a high-quality education are based on evidence.
Erica: Likewise, as new materials and processes make it to commercial reality, we need a workforce prepared to take advantage of those jobs. This requires workforce training (and retraining). Indeed, it’s a big issue that in classical economics, technology-development policies and workforce-training/workforce-development policy are not connected. The two are deeply interlinked.
Maryann: The three topics are related — science creates new knowledge and creative discoveries. The easiest way to transport knowledge is to wrap it up in an individual. A strong educational system is needed to realize the benefits of science. Advanced manufacturing is an application of science.
Maggie: Let’s talk a little about these specific fields and what Clinton has been saying about them. I want to start with advanced manufacturing, partly because it’s the field I know the least about. Erica, what is advanced manufacturing? Can you define that for us? Clinton has talked about it both in the context of factories making new materials and factories working in different ways — but it’s not been really clear what she means, specifically.
Erica: If there is a formal definition, I am not aware of it. That said, in some ways the definition is simple — advanced manufacturing is manufacturing capabilities or areas that are new and forthcoming. It’s how we’re going to make things in the future. What that future will look like is a harder question.
In the short term, “important areas” often cited by the White House and consulting groups include advanced materials, additive manufacturing, continuous manufacturing of pharmaceuticals (rather than batch), biomanufacturing for regenerative medicine, specialized medicine, advanced robotics, smart factories, and internet of things.
What might these words mean in the next 10-20 years?
- Advanced materials: Imagine intelligent clothing, creating plastic shapes with a beam of light, building veins or infant hearts.
- Additive manufacturing: Casually referred to as 3-D printing, but imagine not just in plastics, but also metals, semiconductors, food — changing materials layer by layer.
- Specialized medicine: Custom vaccines and medicine tailored to your genetic makeup.
- Advanced robotics: Imagine little robot modules that reassemble to achieve — jointly with humans — different tasks.
- Smart factories: everything connected with data and real-time analysis.
- Internet of things: All products talk to all other products, and in the longer term, the product tells you why it failed and how to build it better the next time.
In the long term, we might imagine a postal worker being directed over Google Glass how to additively manufacture on your doorstep the sneakers you ordered 5 minutes ago.
But I think the real issue is slightly different. To have jobs, the U.S. must have a comparative advantage — it must be able to produce more efficiently than another nation. Otherwise we have nothing to produce and no jobs.
Maggie: Our economics editor, Ben Casselman, has pointed out recently that manufacturing is on the rise in the U.S., but that manufacturing isn’t creating a commensurate increase in employment. When advanced manufacturing comes up in her speech transcripts, Clinton is often talking about it in the context of creating jobs. But she’s also said that we have to be careful with advanced manufacturing because we don’t want to end up with a system where we’re making a lot, but actually laying more people off. I’m curious what Erica and Maryanne have to say about what policies could balance that. Or is that something policy can control?
Maryann: Absolutely, there can be balance. Policy can incentivize training for the new jobs that are created. Policy can also guarantee that wages are at a level so that families can have good quality of life and in turn buy products and services, creating demand for other industries and jobs for other workers. Policy can also ensure that we have competition within industry sectors.
Elizabeth: It seems that automation will have an important impact on the number and types of jobs that are available, and that this will factor in heavily to what kind of worker training we focus on.
Erica: There is empirical evidence that manufacturing has more jobs and higher-paying jobs than other industrial sectors, and that advanced manufacturing will have these same qualities, possibly more so. [Editor’s note: Some economists have a different take.]
Maggie: And this kind of goes to a question that I know our economics editor and I have had here at FiveThirtyEight: Does it make sense to spend a ton of time promoting advanced manufacturing, or should we focus on improving the jobs that exist in the service sector now?
Erica: We should absolutely not be focusing on improving the jobs that exist right now (whichever the sector.)
Maggie: Wait, wait. Seriously? Then what happens to the people who are stuck in crappy jobs now?
Erica: Thank you for helping me clarify. What drove my response was the speed of technology change. While basic research or science investments can take a long time to create industries, technology change is very rapid. The jobs and the knowledge relevant today are not the same jobs or knowledge [that will be] relevant tomorrow. [If you] set up programs to improve today’s jobs, those programs will be out of date before they are implemented. We must set up programs that prepare workers for the jobs and knowledge needed next month, next year.
Imagine an assessment and training application — accessible anytime, anywhere — that Uber drivers could access on their iPhones with Google Cardboard to train them for the next job needed by the economy by the time their current one had been replaced by a driverless car. I would wager Uber drivers being replaced by driverless cars isn’t that far away. But are our training systems ready to reach the population that has been reached by Uber?
Second, if we focus on the service sector and [we] aren’t making anything, we are putting our national security severely at risk. It’s a nonstarter. Some of the most important advanced manufacturing technologies are materials — new materials. And research suggests increasing the number of manufacturing jobs would have a more immediate impact on jobs and the economy than improving pay and working conditions in many service-sector jobs (where competition abroad suggests the higher wages [in the U.S.] will merely mean we don’t compete).
Maggie: That could factor into some of Clinton’s other policy goals, like installing half a billion solar panels by the end of her first term.
Erica: We know historically that new materials stay locally in their early years (e.g., don’t go to China) and need people on the line, because we don’t know yet how to get good product out the door.
But the future of automation is more likely to look like humans working side-by-side with robots [Editor’s note: clicking the link will download a PDF] (imagine small reconfigurable balls or modules that take different shapes), doing what each does best.
Elizabeth: Interesting! So, it seems like concerns along the lines of “robots will take all our jobs” are exaggerated, then.
Erica: Yes. A robot today can’t even open a bottle cap. [Editor’s note: That may be changing, depending on the type of bottle.] There’s a lot that robots (and artificial intelligence) can’t yet do. The silver bullet is for us to develop machines that can leverage what humans are good at (creativity and problem-solving) and support them in things they’re less good at and don’t want to do (repetitive tasks that are boring, dirty, heavy, etc.). As for repetitive or dirty tasks that would take classic automation — those probably aren’t the ones we want in the first place.
Maryann: It’s an old claim. The truth is that there will be job displacements, but we think we will all be better off. Economic theory dictates compensating the losers, and this is an area that policy needs to address.
Maggie: Can you talk about “compensating the losers”? What would that look like? Education and job training? Or something more structural?
Maryann: Extending unemployment benefits, providing relocation and job-training assistance.
Erica: I agree with Maryann’s comment on job displacement and about policy’s role. In the ’80s, there was a lot of fear about losing jobs to offshoring and robots. In the end, there were more jobs, but more at the top and more at the bottom and fewer in the middle. Particularly fewer middle-class manufacturing jobs.
Elizabeth: I think the “more at the top, more at the bottom” issue is tricky from a political sense, at least when both candidates this year, and the last several years, are fighting to be champions of the middle class, albeit via very different messaging and policy strategies.
Maggie: I want to take a question to Maryann. Clinton said she would increase funding for scientific research, including basic research, with an eye toward “high-risk, high-reward” projects that can create whole new economic fields if they work out well. She’s specifically brought up the Human Genome Project. What do we know for sure about how basic research contributes to the growth of the economy? I mean, I can name lots of examples of curiosity projects that, 30 years down the line, led to something commercializable. But how do you quantify that?
Maryann: Basic research is highly uncertain, which is why government funding is needed. If R&D has a high and relatively certain rate of return, then the private sector has an incentive to step in and do the research. But in the absence of government funding to de-risk new discovery we quickly lose the supply of industrial R&D.
Maggie: Clinton has said that federal funding of research is tiny — less than 1 percent of annual spending — but has big returns. Does what you’re saying here contradict that? Or is it more about how, if you hit the jackpot on some things, like that Human Genome Project, the overall return can be big even if every investment doesn’t have a big return?
Maryann: The wonderful thing about basic research is that we get an increased understanding of the world. It takes a long time to realize benefits. But the investment has been falling. And shifting to NIH and away from other science agencies. So all the evidence is that return on investment in science is large and has a long time horizon.
We are going to see a revolution in personalized medicine and better health as a result of the Human Genome.
Maggie: Does it make sense then, to expect funding of research to be a job creator in a single presidential term or two?
Elizabeth: I think that one thing presidents consider during their time in office is not only what impact they can have in four or eight years, but also what their legacy is. So while investing in something like mapping the human genome (and other projects with long time horizons) may not produce tons of jobs by 2020, it’s definitely something that Clinton might want to point to as part of her longer-lasting impact. But of course, concerns about legacy are different than concerns about getting elected in the first place, and Clinton’s current rhetoric may reflect that.
Erica: If I might add to Maryann’s comment … in relation to your question. If we’re talking about basic research, a presidential term or two is probably not the right time line, but that doesn’t make it any less important. (Ah! Elizabeth has said it better than I could!)
Maryann: Ah! We are so concerned with the short term. But increased funding for science creates jobs, which are needed for the glut of doctoral students we are producing. Immediately, there will be a multiplier effect as lab equipment, computers, etc. are purchased and as scientists spend their salaries. I agree it’s a longer-term legacy. Aren’t we glad the JFK took us to the moon!
Maggie: And, with it, velcro. And Tang!
Maryann: And teflon and better civilian aviation.
Erica: If we’re talking research and development, then it could have impact within the time of a presidential term or two. One professor here at Carnegie Mellon, for example, had an idea, and 10 years later has a very successful award-winning company manufacturing stationary grid batteries. That’s fast for that type of project, but it can happen. Something even closer to development (e.g. not a brand-new idea) could have impact even faster, especially in terms of workforce training matching to new technologies at manufacturing readiness levels.
Maggie: I’ve got some questions for you on the education front, Elizabeth. Clinton has said she wants universal pre-K and better access to STEM classes. What can you tell us about the distribution of resources and STEM classes for Pre-K to 12 across the country? One of the things I’m curious about is how bad the disparities of inputs are. (Inputs are the resources that are important for successful STEM education. We talk a lot about disparities of outcomes, but I’m not sure I’ve seen any measurements from the other side of the equation, like number of computers available at school and what disparities exist there, or age of textbooks, or schools with AP physics. That kind of thing.
Elizabeth: There are definitely disparities in terms of inputs for STEM education at the K-12 level. In terms of inputs, one of the first things that comes to mind is teachers. It’s harder to hire and retain STEM teachers, and this is especially pronounced in schools that predominantly serve low-income students. It’s also the case that there are racial disparities in terms of access to higher-level math and science courses, which is another problem on the “input” side of the equation.
Maggie: I’m also curious how common you all think it is to see a presidential candidate treat education as a science issue? Because as I read through these speeches, I sort of realized that I’m used to hearing about them as though they are two separate things.
Elizabeth: On the question of education as a science issue, the increased focus in recent years on scientific evidence as a basis for education policy may have contributed to the reframing that you mentioned, Maggie, where education is treated as a science issue.
Maryann: One distinction is pre-K-12 education, while the other related to higher education, including community colleges and vocational training programs. The U.S. has a higher-education system that is the envy of the world, but our K-12 pipeline has serious problems. In a world where technology changes quickly, there is a need for every student, and every citizen, to be educated and scientifically literate. It’s a workforce issue.
Maggie: So much of this, from the hiring to the very existence of pre-K, is a local issue, though, right? And while it’s relatively easy to put in federal testing mandates (which is why, you know, the last two presidents have done that), it seems like it would be a lot harder to federally affect the existence of pre-K or whether a STEM-trained teacher can be employed by a poor district. What are some ways that Clinton might be able to achieve those goals?
Elizabeth: This is an issue that is actively being sorted out. The role of the federal government in education has been hotly contested since it took a more activist role in education policy. With pushback to No Child Left Behind culminating in passage of the Every Student Succeeds Act last year, it hasn’t gotten any easier for the executive branch to make the case that the president and/or the federal government should be responsible for how state and local education is delivered to students.
So from a policymaking standpoint, the Department of Education under Secretary John King seems focused on using the rulemaking process to shape implementation of ESSA and to push its policy goals. That could be a strategy that Clinton continues, although King has faced a lot of public criticism for this particular technique, so it’s unclear whether Clinton would want to keep going down that road.
Maggie: So this is kind of a case of Clinton saying, “Hey, I’d like a pony!” and sort of hoping that everybody agrees it’s OK for her to go buy one?
Elizabeth: Ha! Wonderful. I think that’s not quite right, as the federal government does have authority to hold states accountable and certainly has oversight authority.
Maryann: There are policies, like federal grants, that can motivate change!
Elizabeth: It’s more a question of figuring out where the bounds of federal authority lie — for example, currently the federal government, through ESSA, has the authority to mandate testing and require states to have standards, but ESSA is clear that the federal government cannot set those standards.
And that’s a good point, Maryann. The Obama administration used competitive grants (Race to the Top) and later, waivers, to pursue its policy goals.
There was pushback to both of those policies as well, so it’s unclear how easy it would be for Clinton to replicate the Obama Administration’s success there. All that said, approving pilot programs seems like a potentially fruitful route for a Clinton administration to address questions like placing STEM teachers in low-income schools, and this is something that ESSA authorizes funding for, in part to help stimulate state innovation in this area.
Maggie: What I am taking away from this is that, if she’s elected, Clinton is going to be very busy.
Erica: Indeed!
Maggie: Thank you all again for participating in this chat! I really appreciate it.