Monday, March 19, 2012

Science Friday covers STEM 'shortage'

On NPR's Science Friday last week, a panel of senior science administrators, an 8th grade science teacher and a lone economist discussed the US' STEM labor pool. I'd say that the panel was mostly focused on "STEM shortage", with only Dr. Lindsay Lowell from Georgetown presenting actual data about the travails of STEM workers. The hour discussion (in a public forum) was a mishmash of really good points and the typical STEM cliches that have marred the debate (Sputnik, STEM enthusiasm, subtly bashing your political opponents, etc.) But there was a nice discussion by an audience member about the STEM is TE concept:
Audience member: I'm curious about the panel's opinion on advancing all STEM education versus picking winners in a sense. That is, STEM, we lump it all together. But it's very broad, engineering is very different than chemistry and math and whatever.  My impression is that the need in those different industries at the present moment probably differ quite a bit in terms of workforce, and so I'd like to know whether they feel that advancing all STEM education is a fundamentally important goal or we should be targeting certain areas.  
Flatow: That's a good question. Lindsay, you found in your report that there were some areas that were open to having more graduates in it, like the biological sciences, that there were shortages in some areas.  
Dr. Lowell: Well, petroleum engineering is one area where wages are rising and more students are going into it. And that to me is part of the prior question as well. No, I don't think it would be a good idea to target particular disciplines unless it was very clear it was needed, say, nursing in the United States. But the uh, the market really sends signals. We know the cyclical flow of students through the engineering pipeline follows wages. You can talk about upping supply, but it's not really going to get you what you want unless you stimulate demand, and that's what the gentleman next to me has a lot of major ideas as well. My major message to you is that it's a demand-driven system, it's what we've got. So the best way to have people get into the disciplines that are needed is to have the wage lead the way. 
Well said, Dr. Lowell. 


  1. But ... why can't we (government) use this wage data to funnel funding to appropriate areas? If we are monitoring this data, we will know where the "need" is. And, we can shift funding percentages to different fields to mirror the need. It seems like this would be a reasonable/agile strategy that the government can use to somewhat help limit the over-population in underemployment in certain disciplines.

    1. This would not work simply because the "old" good professions expire faster than we can implement changes. Around the Civil War time a professional/craftsman learned a trade and worked in it his entire life. Around 1980s that cycle has shortened to perhaps 15 years. In the last decade the trends and needs change every 3-5 years.
      Even if we implement an immediate change in education (yeah, right...) the first fully educated college graduates will start their first jobs in eight years (high school + college). This would put them two job cycles behind...

      What we need is a credible continuous education system that can re-train highly educated individuals in a short time (~1 year) and does not cost an arm, a leg and the first born. That would be the most efficient safety net for us right now.

      And there is the "other" competition, too. An interesting interview about it with the authors of "Race Against the Machine" aired this Saturday on Innovation Hub at WGBH Boston.

    2. Interesting. I think that what I was getting at was research funding. Should we be "funding" research in areas taking into consideration what current job prospects are? This would cut down the time ... I don't know that it would correct anything as you say that trends change every 3-5 years. And I agree that finding a way to efficiently re-train people would be really important. You would think that someone like DARPA could figure out how to put an advanced re-training center together.

    3. Maybe industry-funded graduate work is what's really needed. Google can fund hundreds of math and computer science Ph.D.s. Dow can fund Ph.D.s in "traditional" fields. Intel can get people schooled in their specializations. Small, related businesses can pool their money. If an industry is hurting, it won't be insisting on educating students for jobs that don't exist.

      Of course, that'll never happen. It would require these companies making gradates happy enough to stay with them to recoup their investments. It would also put some pressure on the universities to keep costs under control.

    4. To educate employees on that scale businesses would need to be able to plan for at least the length of training. With the quarterly business cycle that kind of planning just can't happen.

      Employee retention can be dealt with. One of my former employers had an educational plan where an employee could get even a Harvard MBA funded by the company. The employee had find a position within the company to use the new skills, and to sign a contract that he comitted to staying with the company for five years after graduation. The contract could be bought out be refunding the cost to the employer.

      That was just a few years back, but now it feels like a couple of generations ago...

  2. Well, chemical biology seems like a hot and well-funded area, but as early post suggested, no one wants to hire people trained in this area. Mismatch? How can students read signals that seem completely at odds with 'real' demand?

  3. Petroleum engineering? Well back in the late 70s this was the hot field until oil prices collapsed and the oil companies fired most of their PEs. The number of new PE graduates subsequently collapsed from about 2500 per year to around 200 per year within a decade.

    Same thing happened to aerospace engineering in the early 70s just in time for the post-Sputnik guys to graduate. The guy I was working next to on the Pepsi bottling line during my summer job in 1971 was a serious telemetry expert with a math PhD but no job anymore - he was in his late 40s when his math job disappeared forever.

    So can anyone tell me the next winner 20 years from today?

  4. There are some good examples of re-training and changing specialties. When the SSC closed in 1993 many highly trained particle physicists went to work for hedge funds and investment banks. We can dispute some of the results (remember LTCM collapse in 1998?), but for a while physics was hot in finance.

    What don't know is how to generalize this idea. How to make transferable skills, well, transferable?

  5. I think the best answer is for scientists to have a solid, fundamental and broad education in the basics and rely on the companies to give further training. The second part seems to be what's breaking down... It seems like companies want someone with specialized training already so they'll be ready to step in from day one (and work for peanuts). If this is what they want, I can see the issue.


looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20