SRS Professional Society Workshop

DATA NEEDS FOR ASSESSING THE RELATIONSHIP
BETWEEN GRADUATE EDUCATION AND EVOLVING TRENDS
IN THE S&E LABOR MARKET

SESSION I

SESSION I: KENNETH BROWN, DIVISION DIRECTOR, SCIENCE RESOURCES STUDIES, CHAIR

Session I


AROUND THE TABLE

Brown:
For the first part of our session, we are going to do our "around the table" sharing of the most recent information that professional societies have on the tightness of the labor market. We did this in a less formal way during our last meeting and found it very useful. This time we asked the people who came here from the associations to be prepared to provide a very brief summary of the basic outlook for employment in their fields.

I have three key questions: (1) in recent months, has the market for scientists or engineers in your field gotten better or worse; (2) what is your evidence for that -- is it a new survey or just anecdotal; and (3) what is your outlook for the next year, better or worse.

Rolf Lehming of SRS is taking careful notes on this because we are going to prepare a one- or two-page memo to the Director of the National Science Foundation and to all those present, summarizing what we think is the latest information and the latest outlook on employment. Of course, as is true with our SRS data, and even the professional societies', the surveys are having trouble keeping pace with the employment market which is a rather fast moving target. We have heard different stories about the market for scientists and engineers getting a little bit better. I'd like to see on the basis of what people can contribute today whether that appears to be true or not.

We'll start with Vin O'Neill from the Institute of Electrical and Electronics Engineers, Inc.

O'Neill:
We primarily depend on the Bureau of Labor Statistics for our numbers -- quarterly statistics on employment and unemployment for engineers, electrical and electronics engineers, and computer scientists and systems analysts.

Engineering employment for the most part is doing pretty well. In the first quarter of 1996, unemployment was down to 1.7 percent for engineers. For electrical engineers it was 1.6 and for computer scientists 1.7. Obviously, those percentages are a lot better than was being recorded two or three years ago. In 1992-93 the unemployment rate for engineers was up over 4 percent.

I think the outlook continues to be rosy. Bob Weatherall, at MIT, writing for Engineers, the quarterly publication that is published by the American Association of Engineering Societies, suggests that college recruiting is busy, if not busier, than it was in the good years in the 1980s.

I think people with computer skills are in the greatest demand. We have noticed a substantial increase in the numbers of employed computer scientists and systems analysts over the past 10 years. It doesn't seem to matter where people develop those particular skills or interests or aptitudes -- they are in the highest demand at least as far as engineering students goes.

Brown:
One of the measures people often use in addition to the unemployment rate, is how long does it take new graduates to find a job. I take it you would say that the time it takes for new graduates to get a job is decreasing.

O'Neill:
I think, anecdotally, that particular index is down a little bit, but I don't have hard data on that.

Brown:
It sounds quite consistent with what you just said.

O'Neill:
Yes.

Brown:
Thank you very much. Michael Neuschatz, American Institute of Physics is next.

Neuschatz:
We track graduating Ph.D.s twice -- once just before they graduate or at the end of the academic year in which they are graduated, and the second time about six months out.

I have data from the class of '95 on their plans. However, what I have now for the six months out data is for the class of '94, since the data for the class of '95 are still being collected. There have been some interesting changes in the `94 class from the past.

For the class of '94, six months out, there seems to be a slight decrease in unemployment compared to previous years after several years of increases in unemployment. There is an increase in the number of permanent jobs for those who had jobs and a decrease in the number of temporary jobs. However, we haven't yet analyzed what type of permanent jobs those are, so we aren't sure whether that represents more people going into the traditional type of Ph.D. employment or whether it represents people finding more diverse employment in physics or even farther afield -- perhaps in finance. You have heard stories about people who go very far afield.

Another trend that we have seen for the plans of the class of '95 is that after a dozen years of an increasing proportion taking postdocs, there seems to be a drop-off. One of the questions is whether the drop-off is due to a saturation in the number of postdocs that could be absorbed due to the consistent and continued increase in postdocs, or to the fact that other opportunities are available, so that a larger proportion of the people who wanted to go directly into work, were able to exercise that option. We'll have a clearer idea of what is happening when we finish our analysis in the next couple of months.

Nelson:
Quite apart from the rates of unemployment, what information do you have about degree production?

Neuschatz:
Ph.D. production seems to have leveled off after about 12 years of increases. Also, we have projections based on the number of people coming in. We are projecting that degree production is going to stay at a high plateau until '97 or '98, and then will start turning down rather drastically. We are guessing that we are going to get something like a 25 percent drop, maybe even more, over the course of the next four or five years. Of course, long term projections are difficult to make accurately. However, we have seen a drop of about 26 percent in the number of first-year students over the course of three or four years -- I think '92 was the peak.

Piekarz:
Does that reflect in some way action on the part of departments?

Neuschatz:
That is something we will talk more about later. For the first time we did ask departments about the motivations for the drop, and got a lot of mixed reactions.

Brown:
While the decline in degree production is not necessarily good news, it is consistent with a market adjustment to the recent difficult employment situation in physics.
Now, we will move on to Mary Jordan of the American Chemical Society.

Jordan:
'95 seems to be the worst year the young chemists had coming out in several decades, and I'll talk more about that later. However, '96 looks rather optimistic for us. It started anecdotally -- the telephone calls coming in to the office were much more upbeat after January. I was getting calls from students with multiple offers, wanting advice on which one should they take.

When I first came to ACS last fall, there was a real desperation in every call I seemed to get, and as some of you know, on the survey last year there was a whole lot of anger among last year's graduates about the employment situation.

Based on the phone calls, I had a feeling that things were getting easier, and so I started sticking out my neck pretty early on. While I have indicators that it has gotten much easier for chemistry graduates this year, I won't know until this fall how accurate I am.

At our national meeting this last March, we had many more interviews and many more openings per candidate than we had in the preceding year. The whole scene was much more relaxed. We had an increase in the positions, an increase in the interviews, an increase in posted jobs, and an increase in unposted jobs per candidate. So that was the first indicator. In the last three weeks, we have had our first two regional meetings and we did almost as well as in all nine regional meetings last year in interviews granted and jobs posted. We also saw a lot more multiple openings for jobs posted, i.e., employers looking for several people to fill a given type of job. However, the numbers are small. For the Middle Atlantic regional meeting, in '95, there were eight interviews for 47 candidates. This year, 34 interviews were granted on site, and only two out of the nine employers that posted jobs were interviewing on site.

Syverson:
Mary, when you talk about positions, are these positions for master's degree recipients or doctoral degree recipients?

Jordan:
All degrees. We are just starting this year to track by degree level. We are surveying the employers as to what they wanted and what kinds of jobs.

The other thing is that I think the students, especially graduate students, who are on the Web, have learned all kinds of ways to look for jobs. It was just reported recently that we have had up to 5,000 hits a day on our Web site -- though that's not necessarily 5,000 different people.

Brown:
Thank you. Next, from the American Geological Institute, Nick Claudy.

Claudy:
The recent market for the geosciences is heavily influenced by what is happening in the oil and gas industries. If you saw the article over the weekend in the Post, there is some indication that that market is beginning to recover from its mid-eighties crash. We are just beginning to get a sense of how much that will drive employment in almost every other area. It will be, they think, a gradual and cautious growth. In terms of degree production for the geosciences, that will probably mean a greater increase in master's degrees than in Ph.D.s, which have remained pretty steady over the past few years. The master's remains the degree of choice for the geosciences.

As Federal employment continues to decline, state employment will probably rise. Further, the consulting field, which is this nebulous mass of those we can't identify otherwise will probably also rise. In academia, we think the employment will continue to decline as departments are forced to become smaller and leaner and/or combine with other disciplines.
Our evidence for this is mostly anecdotal. We are woefully behind the curve in terms of gathering evidence on the employment picture. We hope to correct that this fall. We do gather enrollment and degree-granted information, but not information on employment.

Thus, the outlook for the geosciences is generally good. However, so much depends on what will happen in oil and gas and it is a bit too soon to tell what is happening there.

Brown:
Very interesting. We have either positive or at least mildly positive outlooks by everyone so far. I now turn to Jim Maxwell at the Mathematical Society.

Maxwell:
Let me update you on the almost final measure of last fall's unemployment for the 1994 doctorates that we obtained in '95.
We produce a fall number that is always a bit high. We adjust that based on more complete information that we get. The number this year was just a bit above what it was the previous year. It was 10.9 percent this year compared to 10.7 the year before. That is what we call our spring number. That compares with the fall number of 14.7 percent, which was the highest of the series of fall numbers that we have reported since 1970, and this 10.9 percent is the highest spring number.

So the experience of the 1994-95 new doctorates was the worst that we have recorded, and certainly the worst over the last 20 years. We are just now beginning to gather the information that will tell us the experience of the people who were receiving new doctorates from July of '95 through June of '96. So I don't have any hard data.

One indicator that we have is a jobs fair. We call it an employment register that takes place at our January annual meeting. When we look back historically, that produces some numbers that appear to be rather good predictors of the overall job market. The situation we usually track is the ratio of the number of applicants to the number of employers who are interviewing. The ratio was 10 applicants to every employer -- which is about the worse that that has ever been.

However, there are some anecdotal reports that suggest that some departments or some individuals are seeing things a little bit better, but I can't place much reliability on those. We will have to see.

You realize, of course, that the mathematics job market is very heavily academic. About 80 percent of the new doctorates we track who find employment in the U.S. take their first job in academia.

Brown:
Next on our list is Cathy Crocker of the American Statistical Association.

Crocker:
You would think being with the American Statistical Association I would have the best data of anybody here. I probably have the worst data. I think getting statisticians to agree on surveys is very difficult.

What I will tell you is what we do have. We run a publication called Amstat News. That is a monthly publication. We have looked at the number of ads that have been run in Amstat News, and that is running pretty constant. As Jim was just saying, a lot of our employment is academic, so a lot depends upon what the colleges are doing.

We also have a brochure listing schools offering degrees in statistics. The number of schools has remained steady, which has been nice. There is always some talk from different universities about closing a department or adding a department. However, it seems to pretty much balance out in those areas.

Our joint statistical meetings where we run a job placement service are coming up in August. Last year, the numbers for the job placement service, both the number of people wanting to interview applicants and the number of applicants were up.

Anecdotally, I can tell you that the number of phone calls we have been having from undergraduates wanting to know about the job market in statistics has increased. We seem to be getting about 10 percent more phone calls from graduating students who want to find out if they need to get a master's degree to work in statistics.

We are also getting an increased number of phone calls from people wanting scholarship information. It seems that there are students out there who would like to continue their education, by getting their master's and doctorates, but need funding.

Brown:
Cathy, as I look at the pages of Amstat News, I see the ads are very heavily skewed towards biostatistics -- to work for pharmaceutical firms and public health as opposed to say economic statistics. Is that a good perception of what is happening -- a boom in pharmaceuticals and biological testing?

Crocker:
You are definitely right. Biostatistics is a booming area. Lots of the phone calls we get are from students who are very interested in going into that field, as well. Whether it is an interest they have naturally or they are developing because of these ads, I don't know -- but there is a definite trend.

Brown:
Thank you. Steve Heinig from the Federation of American Societies for Experimental Biology.

Heinig:
I work for Dr. Garrison, who sends his apologies. When Dr. Garrison was last here, we talked about our intentions to poll the members of the FASEB societies. We are still working on those plans, but have not made much progress. We have to get permission from the administrations of each of the societies before we can actually contact members. We are continuing to follow through on that, and we will let you know of any plans coming up at that time.

That is all I would like to say.

Brown:
Thank you. We will look forward to hearing about that. Next on my list is Janet Shoemaker of the American Society for Microbiologists who couldn't be here, so let's turn to Jessica Kohout of the American Psychological Association.

Tupek:
She is ill, but she sent a fax. She apologized for not being able to attend, and said that if she had been here, she would say that she has no new hard data on the tightness of the S&E job market. The 1993 data are the latest available. The doctorate employment survey for 1995 is in the field right now. There has been lots of talk, speculation, and jitters. Requests for data on careers and employment appear to be increasing. This may be a function of improved technology, E-mail and Internet, as well as heightened concerns.

Brown:
Next on our list is Cynthia Costello of the American Sociological Association.

Costello:
We, unfortunately, do not have current data on the experience of the most recent Ph.D.s in the labor market, though we hope to be able to collect such data in the near future. However, we did conduct a survey in 1992, that some of my comments are the result of, but first, let me comment on the trends in the production of Ph.D.s in sociology since this is part of the important context for assessing the labor market.

There was an upward trend during the 1970's, peaking at a high of 734 Ph.D.s awarded in 1976, followed by a downward trend in Ph.D. production for sociologists during the eighties. We are now seeing an upward trend again. In 1994, 550 Ph.D.s were awarded.

In terms of the labor market for recent Ph.D.s, when we look at our own ASA employment bulletin, which is primarily an academic listing service, we see fluctuations over the 10-year period from 1982 until 1992. In 1982, 444 jobs were listed. In 1992, the number was up to 748. In 1992, about 500 Ph.D.s were awarded, and as I just mentioned, almost 750 academic jobs were listed.

However, it is important to recognize that our job listing service lists jobs at all ranks, not just entry level jobs for new Ph.D.s. Some of the jobs advertised are never filled, and the pool of academics applying for any one job is certainly broader than simply the Ph.D.s awarded in the previous year.

We do know that many more Ph.D.s are produced each year than there are academic jobs for them to fill. As a result, as has been mentioned for some of the other disciplines, we are seeing an increasing proportion of our members employed in applied settings outside academia.

In 1992, about 80 percent of our members were in academic jobs, whereas, about 20 percent were in applied jobs, spread fairly evenly across the private sector, government, and self-employed sectors. This compares to 10 years earlier, in 1982, when only about 10 percent of sociologists were employed outside academia.

So when we look toward the immediate future, our best guesstimates are that Ph.D. production is likely to remain stable or experience slight increases. Employment in the academic sector is also likely to remain stable. Enrollments in sociology have been up in recent years, but one implication is that we are likely to see an increasing percentage of sociologists seeking employment outside academia.

Brown:
That is an interesting feature, the question about whether the jobs are in academia, in industry or where. I will mention in advance, one of the features of this afternoon's session is our regular mining of the inexhaustible SRS databases to present some nuggets for your entertainment, and it is going to feature some data on where people are finding the jobs after they have been out of graduate school for a while -- what fields and what sectors they gravitate to.

Next is Patsy Evans from the American Anthropology Association.

Evans:
We have had a steady number of Ph.D.s produced for over a decade, about 300 to 400 a year. We are also having a fairly steady number of jobs in academic employment, which is our major source of employment. There are generally about 350 openings that are advertised through our placement services, which is our best handle on that.

However, in the past year and a half, we have lost three departments of anthropology. We are looking at rearrangements in the university setting which will have a negative impact on the academic employment of anthropologists.

We are having a very large increase in the number of bachelor's and master's students. However, they are primarily in the areas of pre-law, pre-med, and, at the master's level, people who are getting degrees in medical anthropology. These people are working in applied settings.

Our association does not have many statistics on the master's students because the universities are not currently tracking them, but we are working on a system to capture that data, because we think that is a growing area for particular types of expertise -- especially in the medical area.

However, the archaeologists and physical anthropologists are at a declining state. There had been a large amount of employment in contract work for archaeologists working with Federal and state construction projects. Because there has been so much of a cutback in those kinds of projects and because of the large amount of data that already exist, there are some cutbacks in the CRM area. So we are not optimistic, but we are not pessimistic.

Brown:
Thank you. Next on my list is Robert Lewis, who represents Graduate Education for Minorities in Engineering.

Lewis:
We are the GEM consortium, and we represent 75 graduate universities and approximately 80-plus R&D employer members. The R&D employer members make a contribution to the consortium, and in return for that, they get to select two fellows each January.

We have been in existence for 20 years. Our primary raison d'^tre, if you will, is to provide graduate engineers at the master's level. What the data show is that the demand with respect to graduate engineers at the master's level tends to mirror the demand with respect to R&D expenditures.

In other words, as R&D expenditures go up, more employer members join the consortium looking for people to work in their industries. As R&D expenditures go down, people drop out, and so there are fewer people who come in. If we take the 1990s, for example, we receive about 800 applications every year for these master's fellowships. We go through a screening process, where the student has to have a 3.0 grade point average, has to be an underrepresented minority as we define it, basically, American Indian, Mexican-American, African-American, Puerto Rican, or other Hispanic.

Over the last six years we have averaged about 600 plus or minus 25 individuals making it through the screening process. We have been selecting out of that pool, approximately 200 plus or minus 25 over the last six years. We are now in a downward trend in terms of the number we can fund. If we believe what people tell us, it will increase next year, but we won't know that until the check comes in between September and December.

Brown:
You make a very important point, that the medium- to long-term outlook for these professions depends upon the spending for R&D. Steve Nelson and I were just talking about the latest compilations by the AAAS of R&D support in the budget. Going out to the year 2002, the projected declines in real terms for non-defense R&D are now somewhere in the middle 20 percent range.

Nelson: That is projections independently by both the Administration and the House Budget Resolution. They arrive at almost the same point by Fiscal Year 2002.

Brown:
That is up from 30 percent.

Bond: Partly because the appropriations level was higher than the budget resolution last year projected, and because of different economic assumptions.

Syverson:
Steve, is that total R&D or Federal R&D?

Nelson:
Federal non-defense R&D.

Syverson:
If the business community increases its R&D, total R&D could, in fact, go up.

Nelson:
Yes.

Lewis:
Two interesting observations I might share with you is that the demand for chemical engineers with respect to minorities, has always outstripped the supply. In other words, every chemical engineer who applies for a master's fellowship in engineering gets selected. On the other hand, there is little demand for civil engineers. We are engaged now in a study that goes back over the past 20 years to see if we can identify correlates with respect to who gets selected, when and why, in order to better understand the relationship to R&D expenditures.

Brown:
Let me turn to Alan Fechter for some summary observations.

Fechter:
Based on the remarks I heard today and what I have been reading from college placement officers and others, there appears to be a slightly more optimistic situation with respect to job opportunities currently than existed in the earlier part of the nineties. However, while that turning around is pretty much true in the aggregate, it's not necessarily true for Ph.D.s or any particular subset of the larger population.

I think the general finding that I have heard, at least from those who were explicit about it, was that the job market for Ph.D.s is reasonably good, but there aren't enough jobs in academia for all of them. That seems to be a message that came through explicitly in some cases and implicitly in other cases.

I think it is important to know, therefore, what kinds of jobs these people are winding up in. Are they in industry? And, if they are in industry, what are they doing?

I am challenging everybody at this table to come up with information that can answer that latter question -- what are they doing? If we don't know what they are doing, we can't make any meaningful statements about training, since training has to be related to what employers use these people to do. I don't know whether just simply knowing whether they are doing R&D versus teaching versus administration is going to be sufficient to give us information about that. That is number one.

Number two, it is clear from what I heard around the table that the master's degree is becoming an increasingly important degree. There, again, we have a lot of ignorance to dispel about what is going on. From what I have heard people say around the table and from what I have been reading, there is more of an emphasis now on applied as opposed to fundamental or theoretical activities in science. Again, that has implications for training.

We need to know why the master's degree is becoming more important. As an economist, I'd say one possible reason for what we are seeing in terms of Ph.D.s versus M.A.s reflects rates of return to investment and training. The payoff to investing in a master's degree would seem to be much more lucrative at the moment than investing in a Ph.D. It doesn't take as much time to get an M.A., so you are out there working sooner, and the M.A. salaries aren't that bad.

The payoff is lower for a Ph.D., where you have to struggle for six to eight years to get your degree, then, for another three to six years and perhaps beyond in a postdoc position at coolie wages before you get into a job that really pays you something.

It is very possible that what we are seeing is a market response to an economic phenomenon, the rate of return to investment in graduate education, that we need to know more about. We certainly need to know what is going on in the employer side of this market. Why are employers -- if that is the case -- more interested in the master's degrees than the Ph.D.s? This may well fly in the face of certain recent reports that have been talking about retooling graduate education at the Ph.D. level to produce more generalized Ph.D.s in less time. This may not be the way to go. Preserving the existing Ph.D. programs is important. They do important work. To try to beef them up by bringing in people who might be better operating at the master's degree level seems to me a very poor set of policy directions to go in. So, there are a lot of interesting things at the master's level that we need to know and I challenge all of you, again, to see what we can do about that.

Again, it is not just where they are working, but what are they doing, what kinds of jobs are these people taking, and why are employers interested in hiring these master's degree candidates. Is the training at the master's degree so adequate that employers are willing to employ master's people rather than Ph.D.s? If that is the case, the nontraditional career opportunities for Ph.D.s are somewhat problematic -- and we need to worry about that.

Finally, and this came out of the discussion about R&D, I like to remind all who are here that R&D is but one independent variable in the demand function for Ph.D.s or, for that matter, any kind of scientists, and Federal R&D is but one small part of R&D.

When we talk about what is going on, it is important to know what is happening to R&D. The fact that R&D in real terms in the government is going down is important to know, However, we really don't know about R&D in the private sector, and, in fact, it is very likely that R&D in the private sector will be closely related to profit rates in the private sector. I just heard on the radio this morning that corporate profits are up in the first quarter of 1996 compared to the last quarter of 1995. That is a very volatile figure that is very heavily influenced by the business cycle, and we don't know what is going to happen to the business cycle. Certainly, nontraditional jobs, as they become more important in the future, will not only be R&D jobs -- and these non-R&D jobs are influenced by factors that we need to know more about.

I guess the bottom line is that things are looking somewhat better this year than they might have looked in more recent past years. The market seems to be stabilizing and perhaps even turning around -- although I think it is a bit too soon to kind of make strong judgments about that.

We need to worry about a lot of things in terms of what Ph.D.s and master's degree students do and what employers are looking for, and we need to know a lot more about those things before we can make any meaningful statement about how many Ph.D.s are going to go out into nontraditional careers and what we should be doing in terms of graduate training in our country.

Brown:
Thank you very much, Alan. I think that was a good summary. Any concluding remarks on this segment of the program?

Kuh:
I am just curious about one thing I haven't heard about from anybody -- the question of international students and international Ph.D.s. Does anybody know or keep track of them, because it is a fairly big public policy question?

Fechter:
I have written an article on that with a colleague that we hope will get published over the summer.
We know from the data from the International Institute for Education (IIE) that visiting scholars from overseas in this country have been dropping over the last two or three years.

Preliminary IIE data on all fields points to a decline in the number of foreign students. However, we don't have figures just for scientists and engineers. We also know from what we have been reading in Science magazine what is going on in the four tigers in Asia where Hong Kong, Taiwan, and Korea are making moves in terms of setting up their own capabilities for graduate education.

In my paper, I have made the provocative statement that the era of large contributions of foreign students to our graduate programs may be past. We may have to assign the immigration issue to the dust bin of history -- which is a challenging, provocative way of putting it, because what is happening is that our hegemony over excellent graduate education is coming to an end, just as our hegemony over superior economic performance, which we enjoyed in the post-World War II period is ending. European countries, as well as Asian countries, may be strong competitors for these students.

Brown:
Thank you very much and thanks to all who contributed. As I said, we will try to summarize this discussion in one or two pages to share with you and others who couldn't be here. I think it is very useful to bring in the very latest information on where the labor market stands.

IDENTIFICATION OF POLICY ISSUES

Now, we move along to our identification of policy issues surrounding graduate education. As Dr. Marrett indicated in her introduction, there are numerous policy issues surrounding graduate education that concern the National Science Foundation -- what are the right programs, what are the right policies, what are the right kinds of support, and, of course, underlying all of this are the questions of what data do we have, what data do we need, what do the data say.

What we have here are three people who are going to discuss these issues in relation to the S&E labor market. Our first speaker is going to be Peter Syverson of the Council of Graduate Schools.

Syverson:
It is a pleasure to be with you this morning. For those of you who are not familiar with CGS, we are the membership association of deans of graduate schools. I have been asked to talk about a couple of issues that we see as important for the labor market.

It is always fun to follow Alan Fechter, because he is very good at talking about some of the things that I want to talk about. There are two issues that I want to deal with this morning. The first has to do with the master's degree. This is an area where we have very little information, although I was very pleased to hear a number of the professional societies, particularly the engineering professional societies, paying attention to the master's degree.

This is the degree, of course, that binds all graduate institutions together, since there are institutions that grant the master's degree as their highest degree. However, the largest master's degree granting institutions are also the largest Ph.D. granting institutions. As Alan pointed out, this is an increasingly important degree for us.

The hottest fields in the United States right now at the graduate level are not at the doctoral level, but at the master's level. They are in the health sciences, in physical therapy, physician's assistants, some of the social work fields, especially those that qualify you to charge Medicaid and Medicare.

I recently had a conversation with some writers from Lingua Franca magazine, who are putting together a book about graduate education and I said to them, I bet when you talk about graduate education, you simply mean doctoral education, and you are going to ignore the master's degree. That is exactly what they intend to do. To them, a graduate student is a doctoral student, and a graduate degree is a doctoral degree. I think we need to move beyond that very narrow view of graduate education and look at the master's degree.
I hope that the Federal government, and the professional societies, as well, in partnership, will look more at the master's degree. So that is my first plea.

Now, I am going to turn around completely and talk for the rest of the time about the doctoral degree -- we all suffer from this same disease.
There are two ways to look at the employment picture. One is the short term, and the other is the long term. We have been talking today about short term. We have been talking this morning about what is happening to students just after they graduate -- and that is very important. However, I think that for graduate deans and for universities, the long term is really very important, as well, because with this seven- to nine-year time to degree for Ph.D.s, it is really the long term that we have to look at, at the graduate level.

If we change curricula and programs now, it will not have an effect for six to nine years out. How long will depend upon the degree field. For chemists who get out in five or six years, it will have a little quicker impact.

I'd like to tell you this morning about a study that is being conducted by the University of California at Berkeley by Maresi Nerad and Joe Cerny. We call it, "The Ph.D. Ten Years Later Study." Some of you may be familiar with this. This is a study with a number of objectives. Primarily, we are looking at individuals who got their doctoral degrees 10 years ago. We want to find out what are they doing, what have their career patterns been, what impact did their graduate education have on their careers, how would they retrospectively evaluate their doctoral programs, and what suggestions can they make to us for the improvement of doctoral education.

The first objective is to determine what are the career paths. Second, we want to learn about reasons for choosing particular career paths. Third, we want to learn about specific employment patterns of women and minority doctoral recipients. I think we are going to have enough women in the sample. Minorities are always a problem, but I think that we will, in some of these fields, have enough minorities to talk about that, and then finally we will look at program characteristics.

One of the things that came out of the research doc study was that size and a number of other program characteristics were related to our quality measure. It will be interesting to see if program characteristics are correlated with what happens in careers. We are going to do a mail survey for "The Ph.D. Ten Years Later Study," with a lot of interviews, supplemented by a series of in-depth interviews with selected doctoral recipients. Maresi Nerad is out in the field right now doing a lot of interviews with individuals and groups of Ph.D. recipients. There are all kinds of wonderful tidbits coming out of these interviews. For example, if you want to find humanists who are employed in nonacademic settings, you go to New York City because that is where the publishing industry is. If you want to find political scientists or economists, you come to Washington.

The first major goal of this study is to develop an understanding of the outcomes of doctoral education in terms of employment. A second, and very key goal for me, is to develop a survey questionnaire and a methodology that we will use not only for the sample of 61 institutions in this study, but also to be something that we can give to other graduate institutions, so they can do the same kind of study for their own students.

The third goal is to provide career information to graduate students and their advisors. We already have short term information in surveys such as the Survey of Earned Doctorates and in a lot of the other work that we have seen. This study will give us a better understanding of how the short term employment information pans out in the long run.

The 61 institutions in the sample were selected to meet minimum size criteria and to reflect the diversity of kinds of graduate institutions. Since a lot of the funding for this study is coming from the University of California system, we have all nine University of California campuses in the study.

One interesting thing about this study for us is how enthusiastic these 61 institutions are about this study. The deans feel this is extraordinarily important for them, because of the accountability issues they face, particularly from state governments. We hope that they will continue to participate in this study, because we are going to ask a lot of them as we go through the study. Who will be surveyed? The five fields are Biochemistry, Electrical Engineering, English, Mathematics and Political Science. Except for electrical engineering, I think they match up pretty well with the study that Rocco is going to talk about at lunch time. The total sample is about 5,000 Ph.D. recipients.

One issue we face with the study of people that got their Ph.D.s 10 years ago is how do you locate them. We have discovered, like many of the other groups, that credit bureaus are one of the best ways to locate people. It is interesting. What we discovered in the course of this study was that universities themselves use credit bureaus. Berkeley, for example, has a site license for Equifax Credit Bureau in their financial aid office, because they need to track people to find them to pay their bills. By the way, the credit bureaus have been extraordinarily useful. As those of you who are in this business know, you can get software that lets you automatically download this information -- for a fee, of course. I got about an 81 percent hit rate from using these commercial services. Obviously, the missing 19 percent are disproportionately the international students -- and we'll see how we do in tracking those down.

We are working on the survey instrument now and will have a meeting in June with our Advisory Committee. We are collecting a variety of surveys that people have used, including, of course, the Survey of Doctoral Recipients. If any of you have really clever ways to follow career paths, we would love to look at your survey, so that we can benefit from your experience, as well. We are going to be mailing the questionnaire in the fall of 1996, analyze the results in 1997, and hopefully, be back to you with some information. The funding from this study, by the way, is coming from the Mellon Foundation. We tell them they are getting a million dollar study for $200,000, which they are.

One primary topic area for the questionnaire is the 10-year career path for these students. This, by the way, is one of the most challenging areas of this study -- how do you efficiently collect information on all of the possible things that people do in 10 years. We don't have a solution to this yet. If someone in this group has a clever way to do that, a matrix or a way to code this, we would love to see it and to use it.

A second important topic area is reasons for change in employment.

A third topic area for the study is their retrospective evaluation of the quality of their doctoral programs. The final topic area is an evaluation of the usefulness of the doctoral program in light of subsequent career experiences.

Let me conclude my presentation on the "The Ph.D. Ten Years Later Study" and reiterate the two issues that we are talking about here. First, the master's degree. COSEPUP recommended really beefing up the master's degree, and as Alan said, maybe we should beef up the master's degree rather than watering down the Ph.D.

Also, I am very interested in the long-term careers of these doctorate recipients. How does the Ph.D. experience really relate to subsequent careers?

Thank you.

Brown:
Steve will be next, but maybe someone has a question they want to ask.

Fechter:
Three questions. I didn't catch the size. I know the cohort is about 5,200. From '82 to what -- what is the end year?

Syverson:
'85, I think.

Fechter:
The second question deals with the issue of getting longitudinal data on career patterns. I wondered whether you have considered focus groups. You might want to do a pilot with a small sample of people before you go out on a whole big survey. The third question is on the evaluation of the value of graduate education. As you are probably aware, the American Economics Association did such a survey five or 10 years ago -- the results of which were published in the Journal of Economic Literature, if I remember correctly. The Chair of that committee was Ann Kruger. I think she is at Stanford. Lee Hansen was involved in much of the implementation of that. They had questions about the value of graduate education. The survey went out not only to the students, but also to faculty to get a sense of whether there are differences in perceptions about what is important with respect to employability, training, and so forth.

Syverson:
That is a good idea. This brought up another point. One of the things that we do want to think about is what do we consider to be a successful outcome. This is a real important issue, because we can't evaluate outcomes without a benchmark on what might be considered a successful outcome. Maybe we will learn from this study what the individuals themselves consider to be a successful outcome. Like the under-employment issue that we talk about every once in a while, that question is going to be a real challenge.

Bond:
How are you going to get at the issues of the quality of the program and its usefulness? Is it going to be open-ended or are you going to structure the questions?

Syverson:
I think we are going to try to structure those questions. We have not yet developed the questions.

Costello:
Certainly, one of the concerns in many disciplines is whether once new Ph.D.s find themselves in a marginal part-time employment situation, they ever get out of it. So, I would think that questions in terms of career history on voluntary versus involuntary part-time employment would be very important.

Fechter:
My question follows on that. I am wondering whether, if you take a cohort of people who graduated in '82 to '85 and track their experience, what will that tell us about today. The information you get will apply to a cohort that is roughly 7 to 10 years since Ph.D.s, and conditions change over the 7 to 10 year period. What can you do for policy purposes with this information?

Syverson:
We have debated that issue. Departments change over a ten year period. It is unclear to me what the immediate policy issues are going to be, but I think this is going to tell the universities a lot about how they have been doing. I am not sure how to do this kind of work other than through longitudinal, retrospective studies.

Brown:
I think we need to move on to Steve Nelson's talk. Thank you very much, Peter.

Nelson:
I should emphasize at the outset that I am a little out of my element here. I don't normally do work in these areas. To the extent that I deal with data and measurement, it is on R&D funding issues rather than people; however, my set of responsibilities at AAAS carries me into a lot of different arenas where I learn something about a fairly broad range of science or technology policy issues.

So, what you are going to be getting is a generalist perspective on specialist issues. I don't know whether what I say here will tell you anything you don't already know, but I thought it might be important to focus a little bit less on specific measures, and more on the general policy environment for these issues.

I begin with the recognition that the most important audiences for the kinds of data that bear on issues of graduate education and the labor markets for scientists and engineers are not necessarily the ones that lend themselves to policy decisions at all. It may be that the most important audiences are employers regardless of the sector, the graduate training institutions themselves, and the professional societies that are represented here. To talk about policy issues means we are restricting our focus to a subset of issues that are considered appropriate to action by some sort of government entity in the interest of the public good, and a lot of the issues we're talking about today are, as you all know, considered to be outside the scope of government and beyond its reach.

But, as with many statistical programs that the government sponsors, it is seen as in the public interest -- and this is essentially a policy stance -- to provide information where the government is not necessarily the primary user, at least in a policy sense. There is an assumption that the relevant sectors will attend to the information and act on it and that the system will, to some degree, be self-correcting.

So, it would appear that a first policy consideration or meta-issue is to provide adequate resources for the study of these issues, and to provide for adequate data collection in the range of sectors that need to be covered, across the range of variables that need to be covered. My impression is that the resources for data collection have not kept pace with the needs that we have as a society. This is particularly so, because the system is in the midst of flux, moving towards a new pattern that we can't entirely foresee.

It appears to me as sort of a non-specialist, that in thinking about the relationship of graduate education to labor market issues, we probably have much better data on the labor market side than on the graduate education side. We also have relatively weak evidence about the relationships between the graduate education and the labor market.

You can point to a whole host of measures that are being collected in each of those two realms. There are legitimate policy concerns within the realm of graduate education, such as issues of entry and admission -- where people come from, the kinds of institutions that are providing training, and the types and range of support for graduate training that Cora mentioned in her introductory remarks. All these things are legitimate policy concerns. On the labor market side, there are legitimate policy concerns with regard to both the supply and demand and the balances between those.

What is interesting, though, is that this country, at least to my knowledge, has not dealt with this kind of situation before. We have dealt with shortages or apparent shortages where the level of capacity in the system we wanted was greater than the apparent supply. Certainly, that was the case in the early cold war years, particularly after Sputnik, but we haven't quite dealt with this kind of situation before. There is more than one important difference. We are dealing with a situation where it doesn't appear that there are any shortages. If anything, the opposite is true -- there is either an oversupply of people or an undersupply of positions. However, there is also a funding issue that is of concern here. Before, when they were dealing with shortages, at least in the common wisdom, there was plenty of money and a willingness to spend it to alleviate that perceived imbalance. Now we have a situation where, if anything, the available funds are tightening rather inexorably.

I thought it might be interesting to approach the set of issues from the viewpoint of an actual policymaker and to contrast that with how you and I might approach them. Our usual procedure is fairly logical. We identify a problem, we try to figure out ways to measure or calibrate or quantify the problem -- usually at the level that we see the problem. We bring that information supposedly, to the attention of people in policymaking positions, and then it is their job to take it seriously and do something about it. However, our assumptions about the efficacy of that model are probably wrong at every juncture in that series of activities.

My sense from talking to people on the Hill and maybe even in the upper reaches of the executive branch is that the views of the policymakers regarding this set of issues are pretty undefined. To the extent that they are aware of it as a problem at all, it is undefined. It is not even on their radar screens.

However, if policymakers were to confront this set of questions, they would want to know some fairly reasonable-sounding things, that it turns out are very, very difficult to get a handle on.

The first question that might occur to a policymaker is how many scientists and engineers does this country need and in what areas? How do we get a handle on that? Do we take an operational definition and read the market to find that answer? To some degree maybe, but it places, I think, extraordinary faith in the hidden hand of something to take that as the solution, or, in fact, to take the operations at a highly decentralized level as the sum total of what you want science and engineering to do for you as a society. So, I think there does need to be some policy guidance in that.

A policymaker's perspective would go according to a sequence that might look something like this. Is there a problem? If so, what is its nature? For whom is it a problem? Is it affecting different sectors differentially? Does it affect my constituents? (This applies not just to Congressional people, but the administration, too, and to particular agencies within the executive branch.) In other words, the bottom line is -- why should I care as a policy actor?

The next logical question -- is there a legitimate policy role for government in trying to correct what we see as a problem, or is it some other sector's responsibility? What should be done? What handles can we get on this? Even if we assume that there is a legitimate government role here, we need to know what leverage do we have in trying to affect the outcome, what kinds of leverage do we have, and who will be affected by that. What resources will be needed -- what will it cost in terms of time, people, effort, over what time period? And the bottom line always is, is it worth the political cost?

Now, in the face of that kind of reasoning, our brave attempts to come up with better data to measure all these things look rather limited. I don't mean to suggest that they are unimportant or trivial. In fact, they are absolutely essential. They are often the only thing that gets you over the credibility threshold with a lot of policymakers in terms of taking these things seriously, so they are extremely important. But, they are set within a much, much broader context, where a lot of other things have to happen for there to be some kind of serious and appropriate action.

It is sort of like the components of an automotive engine. Everything is essential, but nothing is absolutely key.

Another idea that I wanted to suggest is that a lot of the policy issues that affect graduate education and the labor market are really by-products of policies that are focused on other issues. R&D funding levels provide a good example. They are focused on other things, but probably have very great effects on the issues of graduate education and employment for scientists and engineers.

There are two sets of categories that Bill Boesman of CRS has outlined that are relevant here. The first is a set of issues affecting the health of the national science and technology enterprise, ranging from R&D funding, to education in science and engineering, research and training, facilities and equipment, Federal labs, and Federal policies in support of private industry's R&D efforts. The second class of issues involves balances between different things, balances between research and development, between civilian and defense-related work, between Federal and private sector support, between big and small science, between considerations of what Bill calls equity versus excellence.

What I am suggesting is that one of the things that we ought to take a look at collectively is how some of these other policy issues affect the issues we are examining here today. I'm suggesting something analogous to an environmental impact statement -- a labor market or educational impact analysis. I don't know where the specific responsibility for this might lie, but I do think we can all agree on its importance.

Lastly, I will conclude with the observation that we have a system that was built on growth, and it was well adapted to growth. That system continues because of past successes, inertia, and a lot of vested interests in maintaining at least certain pieces of it. However, the environment is different now and probably really different for the foreseeable future in a lot of ways that we have already talked about this morning. So, a lot of the assumptions that were built into that overall structure are going to have to be reexamined.

We are trying to maintain an overall science and technology system in this country that has been extraordinarily effective in a lot of different ways, without causing too much pain in the course of a transition to what will probably be a set of different arrangements.

As I was suggesting earlier, the old order has become unfrozen and we are groping toward a new set of understandings about what roles science and technology needs to play in a policy sense on behalf of the society -- and we aren't there yet. We are trying to maintain institutions and adjust them without an overall picture of exactly where we are going.

Brown:
Thank you, Steve. You have certainly taken on the big issues, and I think that could inspire any amount of discussion that we had time for, but since we're late, let's move on to Charlotte Kuh's presentation, and then if there is time for general discussion afterwards, we can have it.

Our third speaker in this segment is Charlotte Kuh, who is the head of the Office of Science and Engineering Personnel at the National Research Council.

Kuh:
When we discuss policy issues in graduate education, we really have to think about who is deciding what. Steve Nelson discussed the Washington scene. What I would like to talk about first is - who cares, i.e., who has graduate education on their radar screens? Second, what do they know? My talk is in the form of questions that different constituencies may have. I am working under the assumption that later in the day other speakers will talk about what existing data can tell us and what data would we like to have if those data are not quite sufficient.

I also want to say that this is one of the few talks I have ever given where you are not going to see a single number. However, I do have two heuristic indicators that I would like to share with you. One is how many offers my husband's postdoctoral scholar has received. He received four this year, where in previous years the number has been as low as one. So I think things may be looking up. The other heuristic indicator is my systems analyst brother in law's duration in a job. Things also seem to be looking up by that indicator, too. There was a while, when the defense industry was restructuring, that he spent as long as four years in the same job.

Now let us ask, who cares about graduate education? First, there are the participants in it, the students and their families. There are also universities or, more specifically, administrators and faculty.

There are also funders and users. Employers are the users of the products of graduate education. Government is a major funder. We must remember to consider both the Federal government and also state and local governments. State governments have a responsibility for many of the institutions that train our graduate students either at the doctoral of the master's level.

Finally, there is the public. When I think about who the public is, I include Congress. The executive branch has programs for which it has responsibility and it has to know things relevant to the administration of those programs. Congress, representing the public, finds itself deciding what it is we should do, if anything, about this graduate education sector, but once programs are in place their responsibility ends--unless programs are public scandals or unless the necessity arises to cut government budgets.

The rest of my talk addresses the question: what do each of these groups want to know, given their interests in graduate education?

Students. In a way, students are the easiest. When they are applying to graduate programs, they want to know the sort of things we heard in the previous talks this morning -- how long will it take to get a degree? Will I be able to pay for it? Will I have a career when I finish? Will I be able to pay back any indebtedness that I might have? Finally, in the longer run, will I be able to make a reasonable income in a job that uses my training?

We must remember that these questions are all predicated on the assumption that this prospective graduate student loves to do science and is pretty good at it.

One thing that strikes me is that, in many ways, the demand for master's education is sort of a modified echo of the growth in the bachelor holder numbers. If we look at GRE test-takers, the peak demand for people taking the GRE was in the depths of the recession in the late eighties and nineties--people who already held bachelors degrees in science and engineering. In the event, it may have been a very good thing for someone with a bachelor's degree to do, and we are now seeing those people who got master's then being happily employed.

Administrators. What do university administrators have to worry about? Well, they really have to worry about whether there should be graduate programs and what these programs should do. They want to know whether their program contributes to the overall mission of the institution. As resources shrink, that becomes a much more serious question because the need may arise to choose among programs. When 25 percent declines in research funding in the next five years are possible, as the AAAS numbers indicate, the option is no longer available to cut every program back by 1 or 2 percent.

So the questions an administrator asks are: Will the program be of a quality similar to other programs in the institution? To improve quality, one has to ask where the program stands relative to other programs in the institution and to comparable programs outside of it? How large should the program be? Is it likely to bring in research funding?

Graduate education is intimately connected to the teaching mission of the institution. The administrator needs to know whether the teaching assistants in the program are needed to help meet undergraduate course demand. Finally, the administrator wants to know what the demand for institutional funds (for fellowships and scholarships) generated by the program is relative to this demand for teaching assistance.

One important way that the kind of graduate education we talk about differs from professional graduate education is that for doctoral education, the institutions typically help pay for some of the education. It is not entirely student funded. When it is entirely student funded or when students have to pay the full burden of tuition, you would expect the market signals to be considerably more direct.

Faculty. One of the things I found interesting in developing this list of questions is that no one seems to be asking the same questions, which may mean that there will be a problem in creating consistent incentives for anyone who wishes to regulate or alter the system.

Faculty members would certainly want to know how heavy the teaching load is and whether graduate students are available to help with that. They want to know whether research funding is adequate. They want to know about the administrative burden of research funding and whether it is tolerable. I recently heard that negotiation with Federal departments that allocate R&D is being described as a Middle Eastern bazaar these days. That probably implies a heavy burden.

Faculty members want to know if there are enough graduate students available to permit fulfillment of teaching and research obligations? The faculty member often has an awful lot to do with the production process of both undergraduate and graduate education. Those obligations don't have to be met with graduate students. They can be met with postdoctoral students, with technicians, and with adjunct faculty--at least for a while.

Finally, faculty want to know whether their students get good jobs. Here, I disagree with the COSEPUP report. I think that most faculty members involved in graduate education really do care about that. Sometimes it works and sometimes it doesn't, but I think that faculty are often surprised by what happens in the market.

Employers. I take some issue with Alan Fechter's remarks in that I am not sure that specific training is terribly important to employers. Certainly, the COSEPUP report did not find that specific training was all that important to employers. I am sure that if people knew nothing at the end at the end of their graduate education of a substantive nature, employers would be unhappy, but that is not the same as asking for a great deal of specific training.

Having done some recruiting for AT&T, I know a little about what employers are looking for. They are interested in things such as whether a student can communicate, organize work, work in a team, and is informed about the state of the art in his or her field. They want to know about flexibility -- will they catch on to how we do things at this workplace? One of the interesting things about the growth in numbers of Ph.D.s is that the attitude of new Ph.D.s is less and less that they are doing the employer a favor by coming to industry, and more and more one of trying to learn what the environment is like.

Finally, I think employers look at what offers they have to make to get the kind of people they want. They want to know if there are enough people with appropriate training that they don't have to increase salaries excessively to attract good people.

Government. Steve Nelson has already talked to you about the questions of interest to the government. My view is a little bit different. The government wants to know what it is spending on graduate education. I don't think we know the answer to that question. Do we know how many research assistants are funded on federal grants? Do we know how many dollars go directly to graduate education, and not to more general kinds of institutional support? Could we find this out through using existing budgetary accounting techniques? I don't know.

Next is our favorite question -- are we training too many Ph.D.s? What do we mean by that? I think that the Federal policy question has a lot to do with the extent to which we are subsidizing the number of Ph.D.s. We don't care whether lawyers who pay for themselves end up unemployed when they finish law school, since there aren't as many Federal dollars involved in that.

We worry if we are spending Federal dollars to train people who then find themselves in work not related to their training. However, you have to think about the nature of the skills in which people are being trained. You know, a graduate who can reason using Fourier transforms can think about derivatives, including financial derivatives on Wall Street. While that isn't physics, it actually is using problem solving skills learned as a physicist.

How effective are existing funding mechanisms? This is the aid question that the National Science Board was worrying about. Again, if you don't know how many dollars are being spent on graduate education, it is very difficult to say how effective existing funding mechanisms are. It is also worthwhile to talk about the linkage between funding people and the outcomes of interest (time to degree, completion rates, quality of training). Again, I would argue we don't know a whole lot about that. However, funders should want to know if the mix or kinds of aid given to graduate students can be improved.

Are the best students going to graduate school? We want really good people to be scientists and engineers. I would argue that there is not a dichotomy between equity and excellence. I would argue that having diversity in our graduate programs is critical for training the next generation of students. I don't know how we are going to resolve that one because there are still an awful lot of people who feel that there is some dichotomy between equity and excellence and don't understand the importance of a diverse workforce to science.

We also want to understand the linkage between graduate education and research, i.e., is graduate education supporting a vital research enterprise?

State and local governments. Government at this level wants to know about state and local benefits. If a graduate program trains people who get jobs in other states, state governments may have less interest in subsidizing that program than a program that trains people who go to work in the colleges, universities, or the industries in the state. When public university administrators are asked about their mission, it may be that how the university fits into the state economy is a very important thing. Another question university administrators care about is how many graduate students do you need to train undergraduates. Again, there is a direct benefit there, partly because graduate teaching assistants cost less than regular faculty.

What does graduate education cost the state, how can this mixture of dollars per full time equivalent student be untangled and divided between graduate and non-graduate education?

Are there too many publicly funded programs, and what are the benefits of graduate training to the local economy? Often, there are graduate programs that serve specific local employers. Some of the really excellent graduate programs in institutions that are not the ones you immediately think of when you think of research universities, are specifically directed to the training needs of local employers and are often directed to the masters level. An example would be the ceramic programs at the University of Toledo near Dow Corning.

The Public (including Congress). When Steve said graduate education isn't on the radar screen, he's right. Sit next to someone in an airplane and say, " I am really interested in graduate education," and see what happens.

If you can get the public's attention -- and you may not want to, because there are a lot of things that are best not on somebody's radar screen -- then the next question the public wants an answer to is why should we pay taxes to support graduate education? They are likely to say: graduate education produces a rich elite with the best jobs in the economy, why don't they just pay for it?

The public also wants to know what are we paying for the teaching and what are we paying for research. Finally, -- and these questions we don't like to think about -- are we paying to subsidize the education of non-U.S. citizens? Why are foreign students teaching our children and doesn't that make it even harder for our children to learn things like calculus?

Those aren't deep questions, but as resources get smaller, we have to justify every public dollar that we want to continue spending. We are going to have to answer those questions and we are going to have to have the data to support our answers.

Brown:
Thank you very much, Charlotte.

END SESSION I