Recent Professional Society Research

New developments in the production and employment of engineers - Dick Ellis, Engineering Workforce Commission of the American Association of Engineering Societies

Shifts in employment and compensation for electrical and electronics engineers - Vin O'Neill, Institute of Electrical and Electronics Engineers, Inc.,

Employment opportunities for the most recent cohort of doctorates in math sciences - Jim Maxwell, American Mathematical Society/The Mathematical Association of America

Changing Employment Statuses in Chemistry - Mary Jordan, American Chemical Society

Employment Outlook for Microbiologists - Gail Cassall, American Society for Microbiology

The SRS/professional society working group on understanding existing measures of underutilization of individuals with scientific and engineering degrees- Roman Czujko, American Institute of Physics

Plans for a Sloan-supported project on the S&E job market for recent graduates - Catherine Gaddy, Commission on Professionals in Science and Technology

ELLIS: I think most of you are aware of what the Workforce Commission and other bodies at the American Association of Engineering Societies are doing. I'll cover the most recent stuff in a moment.

The bulk of our research and employment intelligence operation at the moment appears first in the new publication, Engineers, which we started this past January. I think among our recent work what would be of greatest interest to this group is our estimation the worldwide production of engineers and computer scientists. We published this in the most recent issue of Engineers with major support from SRS and the Foundation in the form of Jean Johnson's earlier published report on Human Resources in Science and Technology in the Asian Region, and access to information that is in the process now, of being published on human resources for S&E's in Europe.

We found that we had a reasonable basis for the first time to test the workability of other databases on the international production of high-tech people. The result of this was an estimate of the worldwide production of first professional degrees -- which we think is at least a starting point and a significant one for engineers in engineering and math and computer science. We expected the numbers to be large, but they were larger even than we would have guessed.

We looked at a total of 235 countries. Of course, many do not produce people with these degrees, but we still havenational country-specific data on, I would think, virtually every producer with a significant number. And we found that five years ago we were turning out approximately 885,000 engineers annually worldwide plus another 123,000 math and computer science people -- for a total of over a million people a year being added to the system.

Those are formidable numbers. One of the most formidable things about them is that the largest players are in Russia and in China -- places where people have not played in the Western market in the past, but are very definitely doing so now.

So we think these data are interesting, and they are available. We do not make a great deal out of their accuracy. I, in fact, rounded our estimates heavily. But I think they're probably as workable an estimate of international degree production as one could expect to have at this time.

I would observe that similar estimates could be done for the other S&E fields, and we talked informally to Eleanor Babco at CPST about doing just that. I expect that at some point we can give people a sense of how we did the estimates. I will say that the amount of work involved is not all that great. I did it in a few days of work out of my house in Hagerstown.

These numbers contribute to a puzzle we are mulling over at the moment. I will simply call attention to it because I think that it sort of defines what we think is one of the most intriguing current issues. What we seem to have at the moment is record levels of outsourcing of work. We have no data on this, but we have anecdotal reports and information -- and not just complaints. I'm taking into account enthusiastic advertising from outsourcing firms. We have record levels of immigrant labor in the United States, and record levels of U.S. employment abroad.

As of the third quarter of 1995, BLS reports over two million employed engineers in this country -- that is a new record. The market appears to be extremely hot, especially for people with computer skills. We are getting anecdotal reports of people paying signing bonuses, raiding each other's staff and that sort of stuff.

So, clearly, the employment market, at least for engineers, has recovered from the deficit of the business recession. The Bureau of Labor Statistics' judgment that the recent unpleasantness for engineers was a business cycle phenomenon and not a fundamental shift in employment conditions looks like a defensible and reasonable interpretation. We had been concerned that we might be seeing some change in the game itself. At the moment, that does not seem to be the case. However, we think the jury is still out on whether such changes are taking place. There is a continuing weakness in compensation that we suspect is new and we want to see whether the healthy employment levels persist for at least a year before we get real sanguine about them. It's possible that the effects of outsourcing and other changes in who does the work will continue. If so, we could see a return to a more difficult employment market for American citizens.

So, these are the issues that are in front of us. I must say it is fascinating to see all this employment in an era so quickly following the business recession. It is impressive.

The brand-new industry/occupation matrix released last week by BLS also suggests that the strength of the employment market, at least for computer scientists and related kinds of folks, will continue for some time to come. The numbers for people with computer skills are phenomenal. The numbers for everybody else are sluggish. But it adds up to a bright picture for people who can deal with the urge to automate.

The growth, by the way, is not in programming. Programmers, I think, face a fairly dubious future. They may go the way of drafting people. Programming itself is getting automated. The hot markets are for people who can figure out how to make the applications work and to design them, that is, for high-level computer scientists, computer engineers, and especially systems analysts. These are the people who will probably be in demand worldwide in the next ten years -- the period covered by the new projections.

Yesterday we got our final figures on degree production for 1995 -- not 1994 but 1995 -- for engineers. As most of you know, we are swift. The figures are interesting in that at the bachelor's level there is virtually no change. Other stable trends that we have observed for years also continued, including a steady rise in the production of people at the master's level and including small but steady increases in the share of engineering awards that go to members of ethnic minorities and to women.

There is, however, one relatively new development which I think will be of substantial interest to everybody in the room. Last year, for the first time in my memory, we saw a fall in the share of doctoral degrees in engineering awarded to foreign nationals. This trend continued this year. The absolute number of those degrees did not fall -- it went up by an insignificant amount. However, the proportionate share went down.

It seems that the long-term trend of increasing participation in U.S. Ph.D. engineering education by people from abroad may be peaking out. This appears particularly likely since we have also been seeing signs of a flattening in our enrollment data. In 1995, not only has the proportionate share of foreign students gone down again to less than 50 percent but the absolute number has gone down as well. By the way, that is a conservative count of who is a foreign student -- it counts only people here on temporary visas, not those here as permanent residents

We have for two years in a row information suggesting that that long trend in increasing foreign participation in graduate engineering has finally peaked out. We have seen similar plateaus in the past, followed by further growth -- so I wouldn't want to swear on pain of death that the trend won't change again. But at the moment, it looks like there is good reason to believe that we've seen it peak. The information we have seen on the growth of engineering education outside the United States tends to reinforce this impression.

So, it is possible that that issue will diminish, or at least not get worse. That is a mixed result for people here because, while the participation of foreign nationals is, to some extent, a problem and a matter of controversy, it is also a huge market for American engineering educators and educators generally. To the extent that that market shifts back to host countries -- as the people who have been trained here return and start their own engineering programs, for example -- then we see a possibility of work that has sustained the educational establishment in the United States begin to drift back also. That would be a very long-term trend, but we may look back in years in the future and say that, as of this point in the game, we were at the peak.

And that is really all I have to say at the moment.

JOHNSON: In the light of Mary Golladay's summary of the 1994 doctoral data in science and engineering, do you think it's still true that the proportionate share of engineering degrees to foreign students is down in 1995? Could it be that the Chinese students who opted for permanent visa status pull down the numbers from the temporary visas?

ELLIS: I don't know.

JOHNSON: I just looked at the data for 1994 that Mary Golladay brought down. The Taiwan and Korea figures are flat from about 1990 to 1994. So there is no growth, no increase in the number of foreign students coming from Taiwan and Korea. For China, though, there is growth, every year about 400 more students coming in. You won't get that if you look at just the temporary visas that you did for your analysis.

ELLIS: Well, we couldn't make a judgment on that based on our particular data. We do not have information on country of origin in the information reported to us. We are merely reacting to aggregates and longer-term trends. My instincts say that it has peaked -- but that's instincts. We're willing to be corrected.

LEHMING: I'm interested where you stand as an organization or maybe you as an individual on the burgeoning numbers of degrees given in other countries in these fields, particularly in some of the Asian nations that we know are coming on like gangbusters in a variety of areas.

I can see one extreme position that would say since we are economically interdependent and also in a somewhat competitive stance with these countries, we should, therefore, generate a larger number of engineering degrees -- let's leave the science ones aside at the moment -- in order to keep up our economic strength. On the other hand, I can hear somebody say that that would weaken the situation of the existing labor force in those areas and, therefore, we should practice what in some other context has been called birth control.

Where do you come out on that?

ELLIS: The organization for which I toil has no position on these matters. My personal judgment is very mixed. As somebody who made a living for a number of years doing cross-national research for this government, in many ways I'm very pleased to see signs that people from the world around are players. I think that is a good thing, personally.

As a U.S. citizen and somebody who watches employment markets for Americans, I think that it is ominous. It seems to me--I may have mentioned this last spring when we met before, the remark of Andrew Grove(?), the CEO of Intel, that he could purchase highly qualified, completely substitutable doctoral talent in India for 10 cents on the dollar. And many people have looked at that claim since and decided that, if anything, it is a conservative claim, that the present costs are lower than that, and that 10 cents on a dollar is a defensible longer-term cost estimate. And it is not a cost estimate that means that we are doing slave labor kinds of things to the Indians. My impression is that Indians are very happy to work for that kind of income.

I think that we see some signs that employers are increasingly looking at an international market when they recruit, while people seeking work necessarily operate within a more local frame of reference. I do not believe that it is sensible for people in engineering to concentrate exclusively on purely local markets. If you're serious about producing technical work, you are necessarily looking at highly specialized kinds of things. It seems to me the market for those things is national and that people who are seriousabout their careers should look nationally and be willing to relocate. But are people looking globally? I doubt it. Some may, but by and large, they're not. Will it be necessary to do so in the future? I don't know. That is a decision that individuals can make; however, there are certainly organizations operating on that level to supply labor.

I think that these could be signs of things we'll see more of in the future. I don't think anybody knows, but it is fascinating to watch. I also think that it has scary implications for people at less exalted levels of training than the folks we're discussing in this room, but that is another story.

Finally, we see continued downsizing, continued tightening, continued display of a kind of macho approach to industrial management which says that people are relatively expendable by past standards. I think one of the reasons that producers can take those kinds of stands is that they do have access to a very large and talented workforce worldwide. What that will do to the attitudes of Americans who have been used to a somewhat softer kind of a world, I do not know. But I think it's going to be extremely interesting to watch -- and interesting in the sense of a Chinese curse.

LEDERMAN: What I've seen in the written material and statements I've heard from companies seem to indicate that some of the more mundane work such as programming is being outsourced, but the higher level work, like engineering design work, is remaining here. Do you have any information or data on outsourcing by type of engineering function?

ELLIS: My own impression is similar, but it's only an impression. Of course, the lower the level of function you talk about, the more jobs you're talking about. My feeling is that there is going to remain a place for talented Americans, particularly people who can design and arrange for the utilization or the creation of technical kinds of things -- the work that the top engineers, top scientists, and top teachers do. But what I see as being increasingly a province of people outside this country is the intermediate level of work.

If you're designing, for example, a very complex piece of software, you may want somebody who is conversant with the U.S. market to lay out the systems design for how the job will be carved up and turned into programmable modules. Once you have done that, the code for those modules can be developed any place. I also think it does not do to underestimate the ability of people abroad. We are dealing with very energetic and talented people.

VOYTUK: The rosy job market that you painted before for graduates -- do you think that holds true for Ph.D.'s as well? We've been hearing lots of stories about increasing unemployment at the Ph.D level.

BROWN: And does it hold true for some of the other fields of engineering rather than computers?

ELLIS: Well, the BLS data don't give you any clues. I would assume that the outlook remains tougher for people at the doctoral level than has historically been the case.

In looking at engineering unemployment, it is important to remember that a relatively small number of individuals have doctorates and that individuals with doctorates have relatively low rates of unemployment. However, rates are high enough to make people in the field scream. Engineers start to scream about unemployment as soon it goes up over 2 percent -- and with some reason.

The unemployment rate is a snapshot of something that is caused by a constantly rotating population of people. So, even at those low levels, it is perfectly possible that over a period of two or three years, 10 percent of all engineers may have gone through a traumatic experience. When you get up to those kinds of figures, then the odds are pretty good that if you're at a professional society meeting, you're going to run into people who have experienced unemployment whom you know, and that feeds anxiety.

I have no doubt that something like that is happening with many people who have doctorates. My guess is that the data would not be visible beyond the noise in the BLS statistics. I take it, on the intelligence I get from you folks, that it remains something of a problem.

We do know that the market for people with doctorates is highly specific as to training and personal interest. In some cases, people with hot kinds of skills are automatically going to have offers. In other fields, where somebody is more driven by internal interests and outlook, it's probably somewhat chancier.

BROWN: Thank you very much, Dick. Our next speaker is Vin O'Neill.

Shifts in employment and compensation for electrical and electronics engineers - Vin O'Neill, Institute of Electrical and Electronics Engineers, Inc.

O'NEILL: I'm going to provide an overview of what has been going on in terms of employment and compensation for electrical and electronics engineers over the past 10 years or so. It is based on our biennial salary and fringe benefit survey. We survey about 25 percent of our U.S. members every two years, ask them questions about what they're doing and how much they're making and what kind of benefits and other kinds of fringes they may have.

This very first slide is just an introduction to IEEE. We have 315,000 members worldwide currently, and 230,000 of those live and work in the U.S. That 230,000 number is down from about 250,000 three or four years ago, so we are having attrition in our U.S. membership.

The organization that I am affiliated with, IEEE-United States Activities, promotes the professional careers issues and technology policy concerns of our U.S. members. Our salary and fringe benefit survey is done every 2 years, and is based on a sample of somewhere between 20 and 30 percent of our U.S. members.

The second slide is an overview of the personal characteristics of our members. You can see that from 1987 to 1995 the median age has been creeping up from 39 in 1987 to about 43 years old currently. This 1995 figure, I should say, is based on information collected in 1994.

About 95 percent of IEEE members are male and 5 percent are female. We're really not making too much progress in providing equal opportunities for women in electrical engineering, although we are doing better, I think.

As far as citizenship goes for U.S. members, around 89 percent were U.S.-born citizens in 1987. It has dropped down somewhat to 86 percent in 1995. Naturalized citizens crept up a little bit from 7 percent to 8 percent. Foreign nationals has gone up from 4 percent to about 6 percent in 1995.

The question on ethnicity always comes up. As you can see, most of our members -- as I guess is probably true in most of the other engineering societies -- are predominantly white male, 92 percent in 1987, down to about 88 percent currently. The Asian-American population is increasing. Theblack population is holding steady at about 1 percent. The Hispanic population has gone up a little bit from 1 percent before 1990 to 2 percent since. American Indians, as you can see, are a very negligible part of our membership.

As far as degrees are concerned, these are statistics based on the question of what's your highest degree. Ph.D.'s are increasing from 15 percent in 1987 to 20 percent currently. Master's degrees of all kinds increased from 32 percent in 1987 to 38 percent currently. Baccalaureate degrees -- again, this is a function of the fact that more members are going for higher degrees -- declined somewhat from 46 to 41 percent. 2-year degrees and "other degrees" are holding about steady.

This is our profile on employment status. The percent employed full-time in engineering was 84 percent in 1987 and dropped down to 74 percent in 1995. There was a substantial increase in the number of our retired members, from 2 percent in 1987 up to 15 percent in 1995. I think part of this is a function of the aging of the engineering population, but it is also a function of the downsizing that has been going on, particularly on the part of our large corporate employers.

When I started looking at this, I expected to find that the number of self-employed had increased, but it really hasn't changed as a proportion of our total membership.

One question that is frequently asked is where are IEEE members employed. As you can see, it has always been primarily in private business, although the proportion has increased from 59 percent to 65 percent. Government employment is declining somewhat. College or university faculty have held pretty steady at 7 or 8 percent.

I had expected to see a drop in public utilities, because of downsizings and mergers and acquisitions going on within utilities across the country, but the proportion of our membership employed by public utilities has remained about 12 percent.

In terms of size of employers, traditionally IEEE members and electrical engineers have been employed for the most part by large corporations. Employment by small employers is growing somewhat, while there is a substantial decline in terms of percentages employed by major corporate employers. Again, this is a function of downsizing, defense acquisitions, and mergers. Where there used to be 15 or 20 major employers in the defense sector, now there's four or five. And with McDonnell Douglas and Boeing talking about merging, there are going to be even fewer, I think, in the years immediately ahead.

The next slide shows in which kinds of industry or service sectors our members are employed. In 1987 and 1989, aerospace and defense were lumped together. You can see there is a substantial decline in the number of people working in aerospace and, similarly, in defense.

Communications employment has increased from 5 percent to 13 percent. Electronics, as you would expect, has increased from 4 percent to 12 percent. The educational sector, as one of the other speakers mentioned, provides a certain continuity and security that does not necessarily apply elsewhere. Medical electronics and biotechnology has been increasing substantially over the past 10 years, while the utilities industries are pretty much constant.

The next chart shows income statistics by highest degree. These are nominal, as opposed to real numbers; however, you can see the upward trend in salaries.

The question that may come up is what's going on here? I think--and this is my own explanation without really having looked at the data--is that there are substantial numbers of people with less than baccalaureate degrees that started out 15 or 20 years ago who are now moving into higher positions within the field and, accordingly, earning proportionately higher salaries. That's my explanation of the disparity here between the 2-year degree and the baccalaureate degree salary differences.

Another question that comes up is what are the changes in the sources of income. Again, my expectation was that the proportion derived from actual salaries would be declining, while bonuses and various performance incentives would be increasing. But that isn't necessarily happening. Salaries, for people who are employed full-time in their primary area of technical competence, continue to be the primary source of their income.

Another question that is often asked is what about disparities in salaries between men and women? Looking at the differences between men and women with similar characteristics, men are usually more generously compensated than women are, at least within our membership.

So that's an overview of what our salary surveys have shown for general trends for employment and compensation for our members since 1987.

LEHMING: The salary data you just showed, is this for your entire membership or only full-time employees?

O'NEILL: It's for our entire membership.

LEHMING: So, a part-time/full-time differential could play into these data?

O'NEILL: Yes. Also, some people report themselves as retired, but are working part-time.

KRUYTBOSCH: What kind of a response rate do you get on your surveys?

O'NEILL: Dick Ellis at the Engineering Workforce Commission, most recently conducted the survey. Dick, do you recall what the response rate was for the last time?

ELLIS: Well, the sample from the perspective of mailing is 25 to 30 percent of the membership base. They then receive a fraction of those back from the respondents. The actual database is in the neighborhood of 20,000 to 25,000 cases or about 10 percent of the membership. We have looked into non-response bias. I have the sense that it is a reasonably decent take. We have also found after carefully looking at the data that the IEEE people are not a bad stand-in for engineers, generally. They tend to be in the middle of the distribution on a lot of things.

KRUYTBOSCH: For example, did you look at the type of engineering degree they have? What proportion of them are actually electrical engineers? Are there mechanicals, civils?

ELLIS: The great bulk of them are EEs, as are the bulk of all engineers -- you are talking here about the largest engineering field. So, there is not a whole lot of crossover on that. There is very interesting information available on fine-cut distinctions on the type of work done. For example, there is information on various kinds of design and research activities, as opposed to management work or personnel administration, by very detailed technical interests. As far as I know, this information is available nowhere else.

There is also a good deal of information available on such things as fringe benefits and other kinds of non-base salary remuneration. To my knowledge, this is by far and away the largest and the only serious --that is to say, professionally conducted -- compensation survey done by any engineering society. There are some other compensation surveys which are reasonably decent. But they're not done on this kind of scale or with the kind of budget that IEEE is willing to put into this project.

LEDERMAN: Do you have any information on post-degree training? Engineers may require short courses and so on to keep people up to date in new areas and so on. Do you have any data on that and how that works with regard to employment?

O'NEILL: I'm not sure that that kind of information is collected in this particular survey. But there is an educational activities board that is very interested in continuing professional education and is collecting information in that regard.

BROWN: Thank you. Our final speaker for the morning session is Jim Maxwell of the Mathematical Society.

Employment opportunities for the most recent cohort of doctorates in math sciences - Jim Maxwell, American Mathematical Society/The Mathematical Association of America

MAXWELL: I am going to present the data that is collected on an annual cycle by three mathematical societies -- the American Mathematical Society, the Institute of Mathematical Statistics, and the Mathematical Association of America. These surveys are directed by what we've come to know as the Data Committee. It has been engaging in this particular survey since the late 1960s.

Let me describe the survey. Starting in May of each year we mail out forms to about 270 Ph.D.-granting departments of mathematics, Ph.D.-granting departments of statistics and biostatistics, and another group of about 35 departments that we call the applied math departments.

We asked the departments for the names, titles and other employment information on each of the new Ph.D.'s awarded from July 1, 1994 through June 30, 1995. We collect that information over the summer and early fall. In addition, every time we get the name of a new Ph.D., we mail out a somewhat more extensive form to that individual, asking him to confirm what the department has reported and to provide more detail about their employment.

We reported in the AMS's membership publication that 14.7 percent of that Ph.D. cohort that got their degrees between July of last year and June of this year were unemployed as of the fall. Now, that's not really what that number is. As I've described to some of you earlier, we gather information that is a reflection of how many of these new Ph.D.'s have definite employment plans for the fall of 1995. And we gather most of that information during the summer. So, this number overstates the unemployment rate, as it is usually defined. However, it is very useful for trend analysis.

We do count as employed anybody who has part-time employment. We have started, in recent years, to get a clear indication of how many are part-time to get a better grasp on the general underemployment that these people face.

Our number of new doctorates jumped this year compared to the previous year. This year we had 1,226 compared to the 1,070s last year. That's actually the highest number (by about 15), since we've been gathering this data through the AMS.

We had an increase in the number of U.S. citizen new Ph.D.'s, which is about 21 percent over the previous year. The actual number was 567. The number of non-U.S. citizen Ph.D.s was close to the record high that we had a couple of years ago.

Another fact that emerges from our data is that there are almost no African American, Mexican American, Puerto Rican, and other Hispanics among the recent Ph.D.s in the mathematics fields. There were six African American U.S. citizens in our census this year and nine Mexican American, Puerto Rican or other Hispanics. And it's been in that range since we've been gathering the data. There are typically no Native Americans.

U.S. citizen female recipients is up 13.7 percent over the prior year; however, that increase is just a reflection of the general U.S. citizen increase. The 25 percent U.S. citizen women, is in the vicinity of where we have been, plus or minus about one percent during the last three or four years. However, that is much better than it was 15 years ago.

The 14.7 rate that we've been talking about is what we call the first report. It's traditionally out in November or December. Later, we get a considerable amount of additional information from the actual new Ph.D.'s themselves and we find out, of course, that some of them did, in fact, find employment. The revised data is usually 2 or 3 percentage points below the original estimate. However, we only get the forms back from about 55 percent of the individual new doctorates. So, there is another 45 percent that we don't hear from.

We've never had numbers anywhere near as high as the 14.7% in the first report and we can go back to the mid-1970s with a comparable number. We collected some different information in the early 1970s. On the basis of that information, 1974-1975 was on the order of this kind of difficulty.

BROWN: But for the years that you showed there, would you say the method of survey, response rate, and everything else has been identical for every year, so we are sure that that huge sudden blip is pretty real?

MAXWELL: Yes. Yes. We've not made any shifts in the methodology along in here at all. So, yes. We believe it's real.

SUTER: But, Jim, I thought I heard you say you have about a 50 percent response rate.

MAXWELL: No. This number is based primarily on what the departments tell us about their new Ph.D.'s -- their plans for the fall. What we get 50 percent on is the follow up with the individuals. And that is mailed out after we get the information from the department.

SUTER: But what worries me about that is that's a key number for you. And that's a very likely biased number, isn't it?

MAXWELL: I haven't looked to see who the nonrespondents are; however, a large group of those who don't respond are those who take foreign employment.

SUTER: Right.

MAXWELL: And this is based on the whole group.

REGETS: Jim, I guess this is related to Larry's point. Have you compared the consistency of the individual's report with what the department had said about them earlier?

MAXWELL: Yes. That is very consistent. Now, remember that what we see happening is that we hear from the new doctorates themselves later in the summer. Some of them have been able to find something since the department sent in the original report -- which seems to be the primary difference between the departments' report and the individuals'. The department knows what's going on.

REGETS: Are there any real differences in the characteristics based on the department saying they have a job between your respondents and non-respondents?

MAXWELL: I don't know.

SHETTLE: I just wanted to comment that in doing my work on unemployment, I looked at the unemployment rate for those who responded by mail compared to the people who responded later. While there were differences in unemployment rates, they weren't as much as I would have guessed. Making that adjustment changed it for 1991 around one tenth of a percent and didn't even affect it in 1993. So, the improvement on this particular variable of increasing the response rate just wasn't very great. I am, therefore, a little less alarmed about a 50% response rate than I would have been before doing that analysis.

O'NEILL: I have a couple of questions. One, the National Academy of Sciences, in their report last spring suggested that unemployment for Ph.D.'s -- now this may be both engineers and science, including mathematicians -- was high. But they were getting jobs. It was just taking them longer. I'm just wondering if that's one of the factors that you may have found.

My second question is where do most mathematicians usually work -- is it business?

MAXWELL: For 73 percent of the 1995 cohort, first employment was in academia, i.e., of those who found employment, 73 percent were employed in academia. And that number has been as high as 80 percent. It's was closer to 80 percent for the whole period from around 1983 to 1993. Our community is heavily driven by the fact that it's almost exclusively academic. That's reflected in the memberships of the AMS and the MAA and also, I think, the Institute of Mathematical Statistics.

O'NEILL: So the question is whether there are too many mathematicians chasing too few jobs, or is it downsizing in academia? Are the jobs just not --

MAXWELL: This slide gives the distribution of where new doctorates were employed. 73.8 percent were in academia and 26.2 percent in government, business and industry. 31.8 percent of the new doctorates were in Ph.D.-granting departments in the mathematical sciences, including mathematical statistics, biostatistics, and applied math, i.e., they were employed by those departments that we survey. 26 percent were in the Ph.D.-granting departments of mathematics. So a little over a quarter of those who found employment in the U.S. went into Ph.D.-granting mathematics departments.

This past year, I looked at some of the Survey of Doctorate Recipients for individuals who had received doctorates from U.S. institutions over the last 45 years. In the employment sector breakout, it shows almost 70 percent of those with mathematics degrees in the education sector.

BAKER: What are your thoughts about the discrepancy between the dramatic increase in Ph.D. unemployment and the continuing increase in Ph.D. production? Is that a time lag problem? Is that a lack of communication? Is it that people study math because they want to and they don't care about jobs? What are your thoughts about that?

MAXWELL: We had a lot of discussion in the late 1980s -- even within our community -- about the notion that in the coming decade that there would be a shortage of mathematicians. So, there was a general atmosphere that the market ought to be good for the 5 or 6 years from 1988 that it might take to get a doctorate. Enrollments went up, while persistence to the degree apparently remained essentially unchanged. Also, starting in a small way in the early 1980s and picking up later in the decade, we had a large increase in enrollment from mainland China.

I can remember hearing from mathematicians, including the governing boards of the AMS that they were really pleased to have these profoundly bright students coming from mainland China. I remember a couple of them making the comment that the great thing about it is that when they finish their degrees, they'll go back to China. However, they didn't.

LEDERMAN: I have two questions. One, what was the basis for your statement that non-respondents include a high proportion of people who used to work overseas. Two, how do you handle post-docs in your survey?

MAXWELL: They're included.

LEDERMAN: Are they included as full-time or part-time?

MAXWELL: Whether they are full-time or part-time, they are included. True post-docs would be employed full-time. We just can't reach the people who apparently leave the country right after they're degreed.

LEDERMAN: So you're not sure that they're a large proportion of the unemployed.

VOYTUK: There are very few of them, though.

MAXWELL: The ones whose status we don't know were taken out of those percentages. People whose employment plans we don't know anything about are excluded from the analysis.

LEDERMAN: In the 50 percent non-response group?

MAXWELL: That, I'm not sure of.

MAXWELL: Returning to the presentation, you see that when we break out unemployment by gender and citizenship, the males are a little worse off than the females; that is, more of them fail to have definite employment plans. Non-U.S. citizens runs a little above the overall 14.9 rate.

Describing what's happened in the employment market for mathematicians, we started in the fall of 1989-1990 to send out a survey to departments, asking information about the department. We asked about the number of positions for which they had been recruiting in the prior year. As you can see in the graph, the number has been steadily down. We are gathering the data now on 1994-1995.

NEUSCHATZ: Looking at the numbers, am I right that the number of academic openings in, say, 1992-1993 was about the same as the number of Ph.D.'s produced?

MAXWELL: Yes, the number of academic openings in 1992-1993 was, I think, a little above the number of new Ph.D.s produced. Remember, though, this is a measure of all positions under recruitment --a lot of positions under recruitment will be senior positions.

MAXWELL: There was a blip in recruitment in 1991. We think that this is the effect of incentive retirement programs. That is conjecture, but we are pretty sure that's why. We didn't suddenly have the group get older.

CONLON: Does any of your data include computer science Ph.D.'s.?

MAXWELL: No. We collected that into some part of the 1970s, but haven't gathered it in quite some time.

BROWN: Jim, thank you very much. It's time for our lunch break -- we'll see you back at 1 o'clock.

AFTERNOON SESSION

Changing Employment Statuses in Chemistry - Mary Jordan, American Chemical Society

JORDAN: Every day in our office we read article after article about the employment situation -- what some perceive as problems in the sciences. In chemistry, we have good news and we have bad news.

At the last meeting of this group, Mary Funke talked about the different surveys that we have. One of those is the Comprehensive Salary and Employment Survey, and I'm going to use that data today in my talk. Most of what I'll talk about today is chemists, since most of our members are chemists. We also have some chemical engineers and some in other professions.

The 1995 salary survey was for all the members of ACS who were full or associate members under the age of 70 and living in the United States. Of that group, we had over 50,000 respondents.

The first graph shows that, in general, over time, the chemists have followed the national trends at a considerably lower level, until a few years ago when we saw a period when we seemed to be going up at a time national levels were coming down. In 1994, however, we saw a down turn in unemployment levels for chemistry. This 1-year downturn follows the engineering data.

There have been changes in all the employment statuses in chemistry over time. In 1990, almost 94 percent of our members were employed full-time. By 1995, it's down to 88.8 percent, with the biggest drop coming since 1993.

This really follows national trends, but it is very delayed. For about two decades, we've seen those kinds of trends coming into the whole workforce in general. Part of this is the aging of America, and part of this is changing work patterns, e.g., more women in the workforce who work part-time.

Post-docs rose almost continually during the last few years -- and had one of the larger growth rates. Part-time workers also rose, as did the percent not employed and not seeking work. We think that the latter group includes people who quit working but have not claimed retirement status.

The nice thing about the 1995 survey was that it was large enough to look at some of the employment statuses within subgroups. One of the most interesting findings is that, at least for our members, we have a real differential in age unemployment rates. Whereas all the other age groups either evened out or turned down in the last year, the youngest, 20- to 29-year-olds, shot up -- and the largest growth in that group was for the 25- to 29-year-olds.

This increase for 25 to 29 year olds probably has multiple causes. Some is that the older unemployed chemists who got downsized out in the past few years had gotten jobs by 1995. Part of it is immigration.

SYVERSON: Mary, this is fascinating. Are these all chemists?

JORDAN: Yes.

SYVERSON: So, this is regardless of degree attainment?

JORDAN: Right.

ELLIS: I have a fundamental question. You're referring to 1995. It still is 1995.

JORDAN: As of March 1st.

ELLIS: As of March 1st -- so if you round it, it's 1994.

JORDAN: Right.

ELLIS: The BLS engineering trends are for later this year. It's possible that if you look later in 1995, you might see the same thing for chemists that we observed during this year for engineers -- that finally it is tailing off more. Do you have any sense of that?

JORDAN: Well, yes, we do because we have the starting salary survey that we send off in the late summer for the 1995 graduates, and we've just started running that. The preliminary data on that shows that their unemployment rates are falling also across all three degree levels.

JORDAN: So, by age, there is a differential. That's probably a part of that post-doc clog in the pipeline.

Another differential that we can look at is that, like many professional societies, the vast majority of chemists are white males. They're either tweenies -- for those of us who were born between the Depression and baby boomers --or the baby boomers. They're white males and U.S. citizens. For white male citizens, unemployment rates turned down for all degree levels between March 1994 and March 1995. For all three degrees, after a rise, as usual, and supported by SRS's work, the Ph.D. unemployment rate is always below the other two degrees.

Comparing men and women for 1992 through 1995, except for 1994, women have been above and remain above men in ACS in terms of unemployment.

KRUYTBOSCH: Does that include for the women also only those that are looking for work, not those that are not in the labor force?

JORDAN: Right. For groups other than white men, the news isn't quite as good. I must caution you here that sometimes we're working with very small sample sizes that are very volatile at times. Our data show that women are less likely to be working full-time than men. Many more have part-time jobs and more are post-docs. However, women are younger and they're closer to receiving their degree than are men. So this may actually be a harbinger of more women coming into the field rather than of women being stuck in the pipeline.

Citizenship, of course, is the big one everybody wants to know. I based this on people who were not post-docs. Overall, this stands for less than 600 people out of 50,000, leaving about 44,000 chemists. The permanent resident and visa categories leapt up compared to the others. But, together, these are less than 10 percent of the total population that responded.

We included Hispanics of all races as a separate category. So these are non-Hispanic African Americans, American Indians, Asians, and whites, and then Hispanics. We actually have more Hispanics than we do African Americans; Asians, of course, being the largest minority.

As you can see, full-time employment is greatest for American Indians -- which was unlike the previous report, and African Americans had the second highest rate of full-time employment.

SYVERSON: These are all chemists, again.

JORDAN: Yes. So, even though the overall trend was for lower unemployment, some of the subgroups aren't having particularly good luck -- and they have vastly different statuses than the majority of ACS members. Thank you.

TUPEK: Thanks. Our next presenter is Gail Cassall from the American Society for Microbiology. She is looking at the employment outlook for microbiologists.

Employment Outlook for Microbiologists - Gail Cassall, American Society for Microbiology

CASSALL: I first of all wanted to start out by telling you a little bit about the American Society for Microbiology. We are one of the oldest and largest life sciences organizations, having been in existence for over 90 years. Presently, we have over 42,000 members, 20 percent of whom are international. We have a very diverse membership in that we represent members from the microbiological and immunological fields that work in the areas of infectious diseases and immunology, as it applies to diseases of plants, animals, and humans. And, in addition, we have a large membership that represents primarily the pharmaceutical industry in terms of development of anti-infectives, as well as the biotechnology industry, and in the environmental sciences as it pertains to microbiology, bioremediation, and ecology.

We were very eager three, three-and-a-half years ago to begin to look at what the opportunities were in this diverse field of the microbiological, immunological sciences, and commissioned a survey to be done of the employers or potential employers for microbiologists. And what I would like to do today is summarize the high points of that survey.

I should first of all emphasize that we've tried to be very systematic and methodical about designing the survey and targeting the people to be surveyed, so that the actual survey and data accumulation only occurred from a period of May of 1995 through August of 1995. You'll appreciate that what I'm summarizing for you today is hot off the press and hasn't been publicly released. In fact, you're one of the very first groups that will be hearing some of these data.

First, I should point out that because of the diversity of the membership and also the diversity of the potential employers, we wanted from the outset to be certain that this survey took into account that diversity. We've used WESTAT to help us identify appropriate companies.

We also planned for the study by holding focus groups comprised not only of leaders in the field, but also of up and coming individuals in the field in these employment areas. We divided the focus groups into four different sectors. The first was the educational sector, representing both graduate and undergraduate institutions, medical schools, veterinary schools, and even dental schools, which would be potential employers for microbiology and immunological scientists. The second focus group involved people from the industrial sector. The third was from the clinical sector, covering clinical microbiology, infectious diseases, and immunology as it pertains to human medicine. The final group was for the government sector.

We found the focus groups to be very profitable in ensuring that the survey includes the things that are important. As a result of these focus groups, we designed questionnaires to target directly those four specific sectors. We feel this has paid off, because the results are very sector-specific in terms of what the employment prospects actually are.

Following these focus groups, a questionnaire was designed and pretested, using members of our public and scientific affairs board, as well as our manpower committee. In addition, 4,000 members of our association were randomly selected to represent individuals in all four of these sectors. We used these as screeners to try to identify the individuals within those employer departments who would be responsible for hiring potential microbiologists and immunologists.

We used those names to select randomly 500 people in each of the four sectors from those identified as managers or responsible for hirings. These are the individuals who actually received the survey. We had an overall response rate -- and this was consistent amongst the four sectors -- of about 47 to 50 percent. Even though we collected information on non-doctoral individuals, i.e., for individuals with B.S.'s and master's in microbiology, and medical technology degrees, what I'm going to show you today only pertains to individuals holding Ph.D. degrees. Also, note that I'll refer to the combination of microbiologists and immunologists in the rest of the talk as microbiologists, but the data includes both.

We did learn a few things that were surprising to us by looking at each of the different sectors. We were surprised to learn that within the educational sector, the undergraduate institutions constitute by far the largest employer, representing a huge market.

In terms of those employers of microbiologists in the industrial sector, we thought we knew who the potential employers of microbiologists and immunologists were in the industrial sector. We knew they were employed in the biotechnology industry, the pharmaceutical industry, food manufacturing, medical supplies, et cetera. However, by far the largest employer were industries in the "other" category. So, one of the things that we're in the process of doing right now is to actually go back and look at the raw data to try to learn more about who these actual employers are.

In the government sector, we might have predicted that the involvement of microbiologists in state and local health departments would lead to this being the largest sector. However, in fact, the federal agencies were by far the largest with respect to the governmental sector.

When we looked at what people were actually doing in these different sectors, the one category that caused us the most surprise was basic research. Given that most of the Ph.D.'s employed in education were at undergraduate educational institutions, it was not too surprising that among those employed in the education sector fairly few actually claimed they were engaged primarily in basic research. Most selected the dual mission of research and teaching, teaching, or applied research.

Not too surprisingly, individuals in industry were mainly doing applied research. Few claimed that they were doing basic research.

With respect to the clinical and medical sector, it was not surprising that most were involved in diagnostic services, but there was only a very small percentage involved in research at all in a clinical setting. We think that this is a big mistake -- we think there should be more emphasis on research in the clinical environments of Ph.D.'s housed in hospital laboratories. Think of the advances that could be made in diagnostics and etiologies of disease of unknown etiology if there were more research in these areas.

In the federal agencies, you'll notice that most microbiologists were involved in diagnostic services. We would predict that this would probably be primarily FDA, USDA, the Centers for Disease Control, and some perhaps at the National Institutes of Health, and also EPA. Notice that out of the four sectors, the place that most basic research appears to be being done is actually in the government sector. That was very telling to us. Even if you look at applied research and development, here again it's primarily in the government sector.

SYVERSON: Let me ask a basic question here. When you say number of departments, what does that mean? Is that the number of individuals, or is that number of individuals in individuals' departments?

CASSALL: That's number of individual departments that would be potential employers of microbiologists. For example, I'm chairman of the Department of Microbiology at the University of Alabama at Birmingham, and I was one of the educators that was surveyed. But at the same time, we have a hospital clinical microbiology laboratory that may or may not have been surveyed. They would be considered a separate department.

SYVERSON: Okay. And one respondent spoke for each of those departments?

CASSALL: Right -- I would not have dared to speak for the hospital laboratory.

What about the outlook for employment over the next three years? By far, the most positive outlook was in private industry. Up to 57 percent of the managers surveyed were positive about the future potential employment of people trained in the microbiological sciences. In fact, if one extrapolates, they would anticipate hiring approximately 1,800 more doctorates over the next three years. In the educational sector, which I'm showing in this slide, overall, 44 percent had a positive outlook.

I would say that we were surprised with respect to both education and industry. One might argue that all of the discussions about the downsizing of academic health centers would have had an impact on hiring in medical schools. But, after all, there are only 123 medical schools, and we saw that the largest employer of Ph.D.'s in the educational sector are not medical schools but undergraduate educational institutions.

In fact, since I first saw these data in October, I've been tracking advertisements in Science and ASM News, as well as through the Microbiology Chairs Association, and believe it or not, I've been getting two or three ads per month looking for potential faculty members in medical schools and undergraduate institutions -- so it doesn't seem to be out of line.

KRUYTBOSCH: It would be fascinating to get a sense of which hiring departments had a positive outlook versus those that had a negative outlook.

CASSALL: That's actually a very good point -- we have not looked at that. That would be very interesting.

It was not so positive -- in fact, it was almost negative in the clinical sector. Only 23 percent overall were positive. This was predictable based on what we know is going on related to clinical microbiology -- in particular with ownership of hospital laboratories. These are not necessarily being owned by educational institutions anymore, but by companies. Also, there are networks being developed. 54 percent had a negative outlook, and 23 percent that were neither positive nor negative.

If you look at the employment opportunities and attitudes in the government sector, overall 31 percent were positive, 50 percent were negative, and 10 percent were neither positive nor negative. Since the survey was conducted between the months of May and October, we're talking current attitudes here.

What about future hiring within the disciplines that constitute the microbiological sciences? I think we have to be very careful in making blanket statements about future prospects in very broad disciplines like the microbiological, immunological sciences, and perhaps even chemistry and physics.

If you look, for example, at what people are predicting they will be looking for over the next three years in the educational sector, it's primarily molecular biology. Infectious disease, and pathogenesis are also high and general immunology is high. Overall, in the four sectors, molecular biology was by far the most common, which is not too surprising. However, it would represent only about 20 percent of the hiring in the educational sector; 20 percent is not that large, given what we know about the importance of molecular biology.

What about the industrial sector? Molecular biology is certainly the most common, but industrial microbiology--people trained in industrial microbiology, fermentation, and genetics, basic biotechnology principles, were certainly the highest category in industry. Again, this was not a surprise.

What was a surprise, however, to us -- and I think to most people who have training programs for Ph.D.'s in academic health centers -- is that the most common predicted need in terms of medicine and clinical was molecular immunology and clinical immunology. And you'll see this come up again when I give you some data on recent searches.

We would expect that information like this probably will have an impact on training in the graduate programs. I know, at least, it will at my own institution. We have a lot of good immunologists, but we don't have training specifically in clinical immunology.

What about the future hiring in government by specialization? You'll notice molecular biology came in second after all other specializations. Infectious diseases, pathogenesis was still high -- probably again representing CDC and USDA, based on a lot of the changes that have taken place, and also possibly EPA related to water quality screening.

What about hard-to-find specializations in the education sector based on the number of unsuccessful recent searches? In the educational sector, genetics led the list, followed by medical and clinical biology. This didn't come as a surprise to me because over the past 12 months in my own department, we had searches for a geneticist. We had a search at the same time going on for an immunologist, as well as a search going on for a virologist. And, it was the genetics position that we still haven't filled yet because even though we had a good applicant pool, it was not nearly as large as for the other two positions.

SYVERSON: Unsuccessful searches is a fascinating topic, because, of course, you could have hired someone. You had applications for that position in genetics. What an unsuccessful search is, is that the department is looking for a very specific kind of person and will not hire some without those specific qualifications -- is that right?

CASSALL: No. In this case, it was a very broad search, and, interestingly, for the top four candidates we identified, we were competing with four or five other institutions that had offered these same individuals positions, because it's a very highly sought after discipline right now with all the major advances that are being made. We wouldn't call ours necessarily an unsuccessful search -- it's just that we feel like that we would rather wait and make our investment in a really outstanding faculty member, as opposed to an okay faculty member.

SYVERSON: Right.

CASSALL: I couldn't say that that would be the case for others. It's interesting, though. If you look at the industrial sector in terms of hard-to-find specializations, it's antimicrobial chemotherapy, something we would not have predicted two years ago. But as you probably know, there has been a lot of public attention and Congressional attention to the real problems that this country and the world are facing right now with antimicrobial antibiotic resistance. Therefore, the pharmaceutical companies are reinvesting in development of anti-infectives. This probably explains the need.

If you look at the number of people currently employed in education or other sectors in the area of antimicrobial chemotherapy, it's one of the smallest areas. So it makes sense that they would be having a hard time filling these positions.

Lastly, in the clinical medical sector, there was a need for molecular immunology and clinical immunology as hard-to-find specializations. In fact, even though this area represents the smallest number of opportunities over the next three years in the microbiological and immunological sciences, there would appear to be a very clear need for people trained in the area of molecular and clinical immunology -- and overall in immunology. This is something that most of us would not have predicted -- maybe this is because the immunologists are also a highly sought after group by industry.

I'd like to summarize by saying we have just begun to analyze the data and to think about how we will apply and share the data. We think that it will be extremely valuable and we're glad we made the investment. I think it has been very important to look at it by analyzing separately the surveys of the different employment sectors.

We have encouraged the National Institute of Allergy and Infectious Diseases to take a look at their training profile in the different areas of at least the medically oriented microbiological sciences -- which they have done. We are now getting ready to compare that profile with what the job opportunities are in the different sectors within clinical and medical microbiology and immunology.

Thank you.

TUPEK: Before we go into the next session, I think Ken would like to pose a question.

BROWN: I wondered if you could give me some help in writing a one sentence of summary, because we have come to the end of the section of the societies' presentations of their latest data. Suppose, when I leave this meeting and head back to my office, I run into our NSF director, Neal Lane, in the elevator, and he says, "How was your conference?" I'll say, of course, that it was great. Then what if he says to me, "Well, tell me, based on what the professional societies said, is the market for new Ph.D.'s getting any better?" I need a sentence to respond to him. I might say, "Maybe." However, I need something a little longer than "Maybe."

Based on some of the very preliminary indications you reported, there seems to be a hint of a turn-around, some improvement in the employment situation -- but these are not very strong signs. Would that be about the right sentiment? Maybe somebody from an association we haven't heard from would have something stronger one way or the other to say.

CASSALL: You've heard from me, and I suppose there is no need to say this again. But I would be a little bit cautious. It probably depends upon different sub-disciplines within a discipline, as to what the employment prospects really are. I think that's something we all need to keep in mind.

CZUJKO: And by sectors of the economy. We're seeing some hints that industrial employment is starting to turn up. Academe has been flat for decades. However, I don't think we've seen the pain in the government sector yet -- and that one could be ugly.

JORDAN: The government sector concern is very large.

BROWN: Okay. I think the Director already knows that one.

CZUJKO: He may have heard rumors.

BROWN: Any other comments on this?

HEINIG: I was going to say I still don't have a good sense of the situation with the post-docs, i.e., the extended period of time spent in post-docs, at least with life sciences. To what extent does that represent under-employment and to what extent does that represent extra specialization for certain candidates and certain degrees?

ELLIS: I would also chime in favor of caution, particularly if there continues to be some evidence that production is running ahead of increases in employment, because then you're piling up more people every year and adding to the problem.

BROWN: Yes, that's a good point. A lot of the figures on graduates, including those SRS data presented by Mary, indicated a continuing high level of Ph.D. production. And, of course, in terms of the federal sector, it's not only the direct employment in the federal sector. We know that the research budget is trending downwards, so that has to have effects yet to come in the demand for scientists and engineers.

I'm getting more cautious by the moment. Fortunately, I get off at the ninth floor and the Director is on the twelfth. So the doors may shut before I have to really put myself on the spot, but I think that what you've just said has been a good summary of the situation.

TUPEK: Just a couple things before we get started again. The next session is a follow-up to our spring meeting. There are lots of issues that were raised during our spring meeting, and there has been some work going on in between. But before I introduce our next speaker, I want to mention a couple of things.

The proceedings from the spring meeting are available on the World Wide Web through the SRS home page. There are cards out there that give the World Wide Web address. And under "Other Reports," you can find the proceedings from the spring meeting. We hope well in advance of our next meeting, we will have the proceedings from this meeting available. I'm not going to promise you next week, but certainly well in advance of the next time we meet.

One follow-up issue had to do with the under-utilization of S&E personnel, something that we sometimes refer to as the under-employment of scientists and engineers. Roman Czujko from the American Institute of Physics is going to talk about some activities of a working group looking at this concept.

The SRS/professional society working group on understanding existing measures of underutilization of individuals with scientific and engineering degrees - Roman Czujko, American Institute of Physics

CZUJKO: I'm here to give you a quick summary of a meeting we had in a working group on understanding existing measures of under-utilization of individuals with scientific and engineering degrees.

Let me just mention who was on the working group. We had Dick Ellis from the Engineering Society, Catherine Gaddy from CPST, Linda Hardy and Mark Regets from NSF, Jessica Kohout from the Psychological Association, and Michael Neuschatz and myself from the American Institute of Physics.

I'm going to try to summarize the discussions we had. If you think that there is any importance in them, you should thank the other members of the group. If you think I've said something completely stupid, then it's probably my fault in summarizing their thoughts.

The basic question here is: What is under-employment and what is under-utilization? Conversely, what is appropriate Ph.D.-level employment? And we've been talking pretty much exclusively about Ph.D.-level employment.

These are not easy questions. There is no easy answer -- and I'm not going to give you an answer. So, if you were expecting one, you can leave now.

We did come to some agreement on what under-utilization is not -- which I found rather useful. We decided that under-employment is not unemployment and it is not those folks who are under-paid. Those are two concepts and statuses that have reasonably good definitions. We know how to measure these phenomena. We shouldn't toss them in as under-employment or under-utilization.

I think all three concepts--under-employment, unemployment, and under-paid -- can be viewed as measures of the health of the discipline, the vitality of the job market. But we shouldn't consider them as different variations of under-employment.

So what is under-utilization? Well, it's a multifaceted concept, and what we spent several hours discussing was trying to identify some of the dimensions of this concept. What I'm going to show are not exhaustive dimensions. In fact, I don't even think that any of them are necessary or sufficient to qualify a situation as under-employment. But they are, in our opinion, some of the facets of under-utilization.

Part-time employment, when you couldn't find full-time employment and being employed out of field when you couldn't find employment in your field, i.e., involuntarily out of field are dimensions. Note, however, there are a variety of reasons for being out of your field that have to do with normal evolution of your career, et cetera, et cetera. Another dimension of under-employment is the level of professional challenge, the use of principal technical competencies, which is a phrase that Dick Ellis and the engineers used in some studies over 20 years ago. We really liked that phrase. Also, certain jobs are temporary because you couldn't find permanent employment. Finally, there is the potential for career advancement.

There are a variety of problems with under-employment, not the least of which is that each of these dimensions has a subjective component. To some degree, it's relying on the respondent to qualify the status. For example, if I am working part-time but I prefer to be working full-time -- there is no objective measure of that preference.

That one is not so hard to do. We're professional questionnaire designers and we can deal with that one. The one that is much tougher is the expectation of what is appropriate and the fact that those expectations are not static. They change over time. The expectations that employers have of computer competency in 1995 for new hires are very different than they were in 1980. They change as you go through different stages of your career. The expectations of what you ought to be doing change. Also, they change according to whom you ask. Are you asking the individual? Are you asking the community of practice? Are you asking employers? Are you asking impartial observers like ourselves?

You're going to get very different answers, and that's what makes this a really tough nut to crack.

What I'd like to do is share some data the different groups on this working group have collected on different aspects of this. I want to mention a few of the kinds of trends that we have been looking at recently that have to do with the Physics Society for Ph.D. physicists and members of related societies.

We asked people, "Do you think you're under-employed?" We used that as kind of an internal validation. Over 90 percent of the people who say they're in a part-time job because they could not find a full-time job also said they believed that they are under-employed. Powerful.

Among those who can't find in-field employment who would like to, half said they thought they were under-employed. Half did not. For those involuntarily forced out of the field, we've been looking at whether the holder of that position feels it is appropriate Ph.D.-level employment. If the employment is believed to be appropriate, the individual may not feel under-employed when they are involuntarily out-of-field.

I'm not counting in this discussion folks who were out of field for a variety of personal choices.

We found a little bit of a trend by years in the labor force. The younger folks who were out of field are somewhat more likely to say they're under-employed than are the folks who have been out 10 and 20 years. For the latter group, working in the degree field is not even the right question. I think that's one of the reasons why we became intrigued by this notion of principal technical competency which can change as careers evolves.

One of the things we asked people to respond to was a Likert scale for the level of professional talent. We found this to be a very powerful predictor of feeling underemployed. Very few folks who felt that their jobs had a high degree of professional challenge said that they thought they were under-employed. Folks who said that there was very little challenge in their jobs were much more likely to think of themselves as under-employed.

When we did the combination, in-field/out-of-field, with professional challenge, what did we find? We found exactly what you would want us to find. For those in-field who felt professionally challenged, 4 percent said they were under-employed. The bare majority of those who were in-field with very low professional challenge thought of themselves as under-employed. Over 90 percent of those out-of-field involuntarily in positions of low challenge felt they were under-employed. Out-of-field, with high professional challenge, 15 to 20 percent felt under-employed.

So, in-field/out-of-field, is one dimension and the level of professional challenge is another. And then you have this really maddening expectation part that affects the whole. Is full-time employment without challenges under-employment? It varied according to your expectation -- and part of that expectation was whether you were in your field or are you out of your field involuntarily? Those people voluntarily out of field for personal or choice kinds of reasons fell right between the other two groups in terms of feeling under-employed. A very nice set of data.

We also took a look at post-docs, since some folks were asking about post-docs. I'm happy to report that post-docs, whether they were post-docs for one year or six years, overwhelmingly -- 90 percent to 95 percent -- said that there was a high degree of professional challenge in their work -- whether they were marking time or not, they love their work. However, there were some differences about whether they thought of themselves as under-employed. That had to do with a couple of things. We found a temporary/permanent factor. Among the folks who were one year out and post-docs, comparatively few said that they were in this temporary slot because they couldn't find a permanent slot.

When you take a look at people four years out, five years out, six or more years out, much higher percentages said they couldn't find a permanent slot. 70 percent of six-year post-docs said that they were in a temporary position because they couldn't find a permanent one.

When you asked them, Do you think you're under-employed? Of the folks who were post-docs, the answer tends to be no, except for the people who say temporary because I couldn't find permanent.

JORDAN: No matter what time?

CZUJKO: It goes up with time.

CZUJKO: As the proportion of those kinds of people goes up with time, so does their likelihood of saying they are under-employed. So, among the folks who are six years out, 70 percent say they're in a temporary job because they couldn't find a permanent; and of that 70 percent, 70 percent say they're under-employed. You have this profound kind of thing.

Now, having said all that, and just giving you a couple of the trends that we've been seeing, we have a couple of other comments to make. One is that I don't think these things address the self-employed. Self-employed as a status involves a lot of non-Ph.D. kind of work: marketing yourself, building, et cetera, et cetera. If you're doing it because you love it and you love the independence, you overlook it. But what if you're forced into this status? And how does that interact with whether you think of yourself as under-employed?

As we're seeing more and more outsourcing and downsizing, we're assuming that the low end of the market of self-employed is going to be a growing phenomenon. I don't think these are necessarily good measures of whether the self-employed will think of themselves as under-utilized.

Similarly, I don't think that these dimensions are necessarily ones that can help us understand the kind of disruption and uncertainty the happens later in the career -- the 55-year-old who was just laid off. I think there's another set of dimensions that we may have to think about.

The bottom line is -- we were all unanimous in saying that what is needed here is a dedicated project to look at the multifaceted aspects of this concept called under-utilization.

That's it. Does anybody have any questions?

TUPEK: Maybe the answer is no, but I'll ask anyway. Is the group at the point to make some recommendations to change survey instruments, the NSF surveys and possibly some of the professional societies' survey instruments that would better get at some of these concepts?

CZUJKO: I think our recommendation was that there ought to be a project that looked at this issue exclusively, not just modifying a few instruments, but designing a questionnaire that will look at the issues we've raised and perhaps others we hadn't thought of yet.

MITCHELL: When you asked people to say whether they were under-employed, did you provide them a definition?

CZUJKO: No, we did not. That's part of the subjective problem. Overall, we were finding around 14 percent of our Ph.D.'s rated themselves as under-employed, slightly higher among the young and certainly lower among the old.

MITCHELL: Do you feel that that term had meaning for people?

CZUJKO: I think that every one of them had a different definition, yes.

MITCHELL: Did a lot of people not answer the question?

CZUJKO: Comparatively few did not answer the question. By then, we had asked so many under-employment-related questions and they were only one yes/no question away from being done.

NEUSCHATZ: Of the ones that did answer that they were under-employed, something like three-quarters also indicated one of the dimensions we mentioned. So people did have some reference in mind when they were doing that.

CZUJKO: Yes, these dimensions kept showing up in a lot of places. The part-timer who can't find full-time employment, the temporary worker who can't find permanent employment, etc.

JORDAN: So, you were asking all different statuses, fully employed and --

CZUJKO: Yes.

SHETTLE: You said that your list doesn't address the self-employed.

CZUJKO: I don't think it does.

SHETTLE: And that's something I've been wondering about. We often kind of sweep them under the rug. Is that because you didn't ask them the questions related to that? You could ask the self-employed whether they are self-employed -- because they want to be or because they can't find other employment. I know people who call themselves self-employed, but really are unemployed or part-time employed. So I think it's potentially a really hidden category in there.

CZUJKO: Yes, and we're seeing increases, second-career folks who have been laid off, rehired by the same company in a temporary status as consultants. It's complex, very complex.

NEUSCHATZ: One of the issues we could have asked was simply to say: Are you self-employed out of personal preference -- rising above the job market, or are you self-employed because you couldn't find regular employment.

ELLIS: Yes, this is in response to Alan's question in a bit more detail. As somebody who was in on the discussion, I had the distinct sense that this is one of these issues that just raises more questions the further you poke into it. I think the committee would be very hard-pressed to think of anything compact enough to make sense in terms of modifying the instrument you use. Before you can do that, my feeling was that you need a more organized effort, a more concentrated attack on the problem to see if it is possible to winnow it down to something that can be plugged in as a workable, smaller, and more doable measure. I don't think we really have a sense of what small subset of items could be used that would give you a workable return on the investment at this point. There's too much going on.

JORDAN: Like creating an index?

CZUJKO: Yes, we might have more questions that predict well enough, but I think that there are probably 15 questions that need to be asked.

JORDAN: But, essentially, you have no concrete outcome.

CZUJKO: Yes. We need to decide what it is we're asking people about, anyway.

ELLIS: It almost makes one wonder if there would be some design that might be factor-analytic, almost like looking at a battery of things and hope that it weeds itself down through some statistical magic, because there's clearly a lot going on. What we don't know is how much redundancy and overlap is there in all of this.

NEUSCHATZ: But, you know, there was a very strong feeling that at this level, especially the Ph.D. level, it's very, very important because unemployment tends to be a very low base factor. A lot of the distress that people feel doesn't necessarily surface in unemployment, but may surface much, much more strongly in under-employment.

JORDAN: But all of the stress research shows that it's autonomy that Ph.D.'s go after and not, you know, some of the things that you're -- that that is what keeps them happy, autonomy in the workplace.

ELLIS: I think we ought to emphasize we're not talking here just about keeping people happy. The under-employment thing also has enormous implications for the effective utilization of talent from the point of view of tax money and all that sort of stuff. Looking at it from the point of view of engineers, we get a lot of people who feel like they're not very effectively utilized but who are gainfully employed and that has horrendous implications for the spending of particular tax dollars. We did a utilization study years ago that suggested that defense industries were particularly likely to yield this kind of outcome. What does that tell you about what we're doing with our money?

There are all sorts of heavy-duty implications of this question. That's another reason why I think we tended to see this as something that maybe merits a hard look -- more than what a committee can do.

SYVERSON: Roman, you've used the term "under-employment" and then "under-utilization" interchangeably. Is this thesame term? I mean, under-employment sounds like my perception of my employment. Under-utilization sounds like something Dick was talking about, which is the sort of society's feeling about under-employment -- is that sort of the way we ought to be thinking about these terms?

CZUJKO: The very term, and what are we trying to find, this is the essential question. I think the conclusion we came to is that there is no one dimension, there is no one thing that is under-employment or under-utilization. There is probably no one point of view. And we even grappled with the possibility that if there's six different facets to this, maybe we ought to have six different terms? I don't know where we're going to go with this.

SYVERSON: So what do you think about the NSF measure of under-employment now?

BROWN: Let me say one thing about that. In the edition of Science Indicators that's going to be published in a month or so, we decided not even to use the term "under-employment," because any time it appeared in the draft in any form whatsoever, some reviewer, often a scientist, would say, "What does this mean? You're talking about the physicist earning $100,000 or a million dollars on Wall Street, he's under-employed? Obviously that's not under-employed." So, we decided not to even use to use the term because it has so many mysterious connotations for the reader. Instead, we decided to be more specific and say a certain group of people responded this on such-and-such a questionnaire. So that's the current NSF approach. Under-employment gets swept under the rug whenever possible.

CZUJKO: Despite the fact that this is definitely a nebulous concept and we asked the question somewhat differently than NSF did on their study, we got almost the same data. If you just look at the first two things -- the part-time who can't find full-time employment, and out-of-field because they couldn't find in-field employment, the difference was that only half of the latter thought of themselves as under-employed. And then there were lots of other folks who thought of themselves as under-employed who didn't fit either of those two categories.

VOYTUK: There's another factor that has to be brought in, I think, and that is what the career perception of individuals happens to be. If they feel they should be a university professor when they get their Ph.D. and they're not, then they may feel that they're under-employed no matter what they're doing, assuming they still have that perception.

CZUJKO: Sure.

VOYTUK: I think in some sense this is a time period when people are going through readjustment and maybe in ten years many of the people in the scientific fields will be more like those in the social sciences or in the humanities and go into areas which they didn't even think of in the beginning.

CZUJKO: One of the problems with the database that we used at AIP is that it's based on society membership, and we are grossly under-represented in terms of folks who are out of field. Presumably, it's overwhelmingly folks who are involuntarily out of field.

COSTELLO: I was very interested in your presentation, and it strikes me that, in addition to thinking through how one taps individuals, a key part of this research agenda is looking at the changes in the labor market, as you alluded to. If the Bureau of Labor Statistics fields a supplement in 1995 on temporary and part-time work, that would be one potential avenue for looking at engineers and scientists in terms of increasing proportions of part-time and involuntary part-time workers, as well as independent contractors, the category that is increasing over time.

CZUJKO: Yes, part-time employment in particular seems to be a very powerful indicator of this self-perception of under-employment. The temporary because they couldn't find permanent employment seems to be one of two categories: long-term post-docs and the part-timers. We haven't found a whole lot of other stuff going on among those with temporary employment. But I haven't talked about career advancement, which is another issue -- rotating temporary jobs that are also dead-end, some of the things that the mathematicians have been complaining about -- folks who get hired to teach for a year and get paid by the course. They have a one-year contract and then they're out of there because they have so many other people to pick from. You can just keep doing it. I think there are some issues related to career advancement under certain circumstances. It's a complicated situation.

TUPEK: Thanks very much, Roman. Catherine Gaddy from CPST is going to talk about plans for a Sloan-supported projected on the S&E market for recent graduates.

Plans for a Sloan-supported project on the S&E job market for recent graduates - Catherine Gaddy, Commission on Professionals in Science and Technology

GADDY: I'm going to tell you about a project that we're just getting started -- we had a kick-off meeting three weeks ago. So far, I'm funded by the Sloan Foundation, and I'm thinking of some other people to perhaps join us.

What we'd like to try and do is fill the gap that seems to exist between the time that some of the federal data comes out and the time that some of the students graduating would like to have the data. Much of the data from NSF and NRC and other federal sources takes considerable time to produce, because of the scope of the data collection and the need to get high response rates.

We would like to take advantage of the society data that is collected -- the smaller scope in a shorter time frame -- and responsibly communicate that to students and to faculty and to others. That is our goal -- to fill that gap in a responsible way. At a minimum, what we'd like to do is have some data for this year's doctoral S&E graduates to tell them what happened to last year's graduates. That's all -- our bottom line is really pretty easy. After that it gets complicated.

This is what we come back to, when we're feeling overwhelmed. Are we doing bachelor's and master's level? Are we doing every S&E field? Are we just doing broad fields? I keep coming back to this bottom line.

So far, we have a modest team to get it started -- most of whom are sitting in this room. We've got six fields represented and our advisory group is growing. We have a few more people we need to invite.

You might say, well, Cathy, why did you pick those fields and not mine or some others? First of all, we worked with societies that already had some survey experience getting data from the graduates, and these six fields had actually provided data to two earlier phases of Sloan projects. So we knew that these folks had rolled their sleeves up and wrestled with some of these issues of collecting data.

We wanted to start with a group that was manageable and affordable and could fit in a room at one time. We also wanted a group of people who liked each other and maybe could agree on a few questions that we might all ask at about thesame time. Sloan has encouraged us to broaden the project to other fields, and we're wrestling now with how do we do that without it becoming overwhelming.

The first task -- which was Roman Czujko's good idea -- was that before we launch into this project, we should compile existing data from the societies to see what we've got and then ask our potential consumers what they want -- what sorts of information would be helpful to graduate students. Maybe we can guess what they would want, but we want to spend a little bit of our research on asking them what they wanted, when they want it, how much they'd be willing to pay for it, maybe -- some questions like that.

We're just now scratching the surface on thinking about how to do this. Some of the societies are going to do focus groups. Some are going to do electronic surveys. Some of them are going to do paper-and-pencil surveys. We don't feel for this task we have to be very rigid and structured. We like to be kind of creative and just see what sorts of information is desired out there.

SYVERSON: So your decision is that you'll focus on the graduate student?

GADDY: When we get overwhelmed, we focus on the graduate student. In calmer moments, we say this data will also be used with the faculty and policy makers.

SYVERSON: I was thinking of the faculty, because they are uninformed and misinformed -- perhaps even more so than the student.

GADDY: I think of the typical faculty members nowadays who are writing six or seven grant proposals with the hope of getting one funded and updating their curricula to include more participative activities and maybe having a life. Then I think -- are they going to keep up with the job market at no extra charge?

SYVERSON: Well, they're being asked to do that by their students. Their students want to know.

GADDY: In fact, we are going to talk to some faculty about what kind of data they would find of interest. We expect there to be a reasonable amount of overlap with what students want, but we're trying to stay focused.

We've also had a number of people ask us about data for policy makers. We don't always know who they are, but we're trying to come up with an operational definition of who is a policy maker. We're going to talk to university administrators, agencies that fund research by scientists and engineers, and journalists; however, we're primarily interested in the students.

The second task, which we'll probably start on late next spring, is to get these six fields together around the table and see if we can agree on some common denominators -- sets of variables that we'd like to capture, things we would like to know about.

Then, we'll do some additional data collection. These six fields already have survey apparatus in place. We would like to influence what they collect. For the professional societies that don't have as much experience or as much expertise as the core group, we'd like to encourage them so that we can level things out, e.g., we'd like to get everybody's response rates up to some minimum.

The actual data collection won't happen until late 1996 and early 1997 in the normal cycles that the societies would be going out to collect data.

Then we're going to pull it all together and start to develop a consistent format for presenting it. One of the things that I know frustrates some of you at these meetings is that you know we're dealing with varying response rates, varying representative mix, varying methods of data collection. So, we'd like to strive for a little more consistency, at least for a subset of the questions.

To disseminate the information, we have the good luck of affiliations with a number of people who are already doing things on the Internet. For example, AAAS has a system that just went up called Next Wave, which you can get to at aaas.org on the World Wide Web. And they're already putting qualitative data up, and we would like to use that as one existing resource rather than create another home page. A second option is that the National Research Council is putting out some resources with a slightly different angle and take on things. Another option is working with NSF staff, to team up with their data. I'd like ultimately to do some calibration of our data and their data, because after two or three years, we'll have a running database going and then we can look at some sort of calibration of that.

So we're going to look for ways to disseminate data, using some existing mechanisms.

This project got kicked off in the last quarter, but we had a lot of planning and transactions to take care of before we could start working on the actual project. If we can get the project in a self-supporting format, we want to have an annual schedule, so that it becomes predictable and consistent.

We've started looking at the surveys members of our group currently do and are sorting different variables and questions into categories, then we will start paring the list down to what we really think are the essentials to meet our bottom line.

In addition to information from graduates, we also want to get information from employers of graduates. We have very modest expectations for this part of the effort. Hearing about that recent study of microbiologists, you can see how complex it can get. We are going to try to get some very modest data from the employers of the graduates, realizing that projections on this kind of data can be kind of fluky. We will focus more on what employers did during the current year.

Finally, it would be good, ultimately, to have some recent employment indexes that would help capture some of these variables.

GADDY: We would like your help on some issues. First, we would like to involve other fields and we are kind of scratching our heads about how best to do this without it being logistically too complicated.

We have the funding from Sloan to fund the six fields. Since they told me we probably pushed our luck with the amount of money we asked for, they are unlikely to fund more than the six fields. Our hope was that if we could get this thing rolling and keep other societies informed of what we're doing then perhaps they would also want to make similar efforts.

If you are interested in being kept posted, give Eleanor Babco or me your card and we will make sure we do that. We are contacting some societies now but I'm sure we will miss some, especially in the life sciences. It's really hard for us to keep track of all the different societies, especially the "boutique" societies.

LEHMING: You talk about graduates, are you only talking about Ph.D. graduates?

GADDY: Yes. The Geo scientists would like to do some master's level folks; however, the psychologists don't even want to talk about anybody other than doctoral. So, at a minimum we're going to do docs, and if other folks want to do other levels, that's fine.

LEHMING: Why focus on people who will graduate this year? You're talking about five-to-seven year to complete the doctorate, on average, and probably longer in many cases.

GADDY: That bottom line is to help keep us focused when we start trying to do too much. If, however, the project has some ripple effects on influencing career decision, even making it back to the undergraduate or high school levels, that is fine. However, we are just trying not to diffuse our scope by making it career planning for everybody. We are trying to keep it focused so at least we can let this year's folks know what happened to last year's folks.

ELLIS: To be literal then, you want to hand people information before they graduate -- assuming conventional May/June graduations -- about what they're likely to walk into based on the experience of the previous year's cohort, right?

GADDY: That's the absolute minimum.

ELLIS: I think that is probably impossible to achieve in a meaningful way, given most forms of information collection, analysis, and dissemination. Among other things, most people, by the time they get to the graduation point, will have been grappling with those decisions for some months.

Aren't you really going out on a limb, in effect? Wouldn't it be equally reasonable and less susceptible of misunderstanding to just indicate you are setting up an intelligence system which would be useful? It seems to me that people who really need it will be the people who are at least one year removed from graduation, because intelligent planning for a career doesn't start in May.

GADDY: Of course it doesn't. This goal is really to remind us that we want the data to be no staler than a year.

ELLIS: I can understand that but it seems to me the way you are presenting it you are laying yourself wide open for a lot of people's criticism.

KRUYTBOSCH: It would be interesting to look at individuals who get jobs prior to getting their degrees. In my own case, I got a job on the strength of having completed my comps but not having finished my dissertation. The expectation was that if I didn't finish the degree, my employment would be terminated. However, I know this information is not available from the data.

GADDY: Again, our goal is really to just make sure that at least we have data for this year's graduates about last year's. That's the minimum to remind us about the periodicity of it.

COSTELLO: I was just going to say that an easy way to make sure that the information about last year's graduates get to the whole set of graduate students would be to disseminate it to graduate advisors. Having this information available and having strong encouragement for graduate advisors to make it available to their entering, first-year Ph.D. students as well as the more advanced students would be good.

GADDY: Definitely. And perhaps the way I stated the bottom line is not the best way to communicate it but it keeps me focused on this issue of the freshness of the data.

We are certainly not going to only let somebody graduating peek at it. It will be available on the Internet, but it reminds me that, as a minimum, I ought to tell students to call us whether they have got a job lined up or not.

MITCHELL: This may be beating a dead horse, but why don't you just take out the words, "this year's graduates"? Because that's where I kind of stumbled too. If you are graduating this year it is almost too likely you have made your decisions. You have already decided to pursue the degree, to pay the money and to incur the debt. Couldn't it just say graduate students?

GADDY: That's fine.

MITCHELL: I think the most likely time that people are going to use this is at the point where they are making the decision whether to pursue a graduate program.

GADDY: But, Susan, the way the job market is changing now, I think it'd be irresponsible to tell somebody who is deciding on a graduate degree what they can expect when they graduate.

MITCHELL: Aren't you just giving them data?

GADDY: Right. But I would encourage them to look at some of the long-term trends. They are not going to be walking into the market place for six or seven or maybe eight years. And they will need to look at even longer term trends to provide career counseling for somebody who is in high school or doing undergraduate work. We are trying to focus on some of the disgruntled folks in the young scientists network. That group of people need a smaller time window on data.

I hear what you're saying. However, if I just said year-old data it just doesn't evoke the same image to me as thinking of a responsibility to folks that are soon to graduate and want to have some data in their hands. Does that make sense?

MITCHELL: Yes, but did you find that by going further, extending your time line and looking at careers over a lifetime you could, in fact, provide some information such as the likely characteristics of your first job; five years from now, what graduates are doing; 20 years from now you might expect a job that looks like this? Did you find that that was just too formidable a task?

GADDY: Oh, definitely. Most of these fields do annual surveys of their graduates and some of them survey employers. We are trying to get it altogether, make it more consistent, get it out there. That's a fascinating topic but that is just way out of our scope.

SYVERSON: It may be difficult for students who are juniors and seniors in college to interpret the data. We want them to look at the long-term trends. We don't want them to look at each year's statistics and say, well, I'm not going to go into physics because this year 14 percent couldn't find a job in June. That's the wrong way to approach this.

So, I like your approach that you are really only trying to help the people coming into the job market, not people coming into the doctoral pipeline. That is a whole different set of questions. I hope some day we will be able to address that question.

GADDY: Oh, it is -- it is much more complex.

SYVERSON: That involves projecting what's going to happen eight years from now.

ELLIS: Would there be a possible leakage of information after they --

ROSCO: I was going to say it might help graduate students planning to enter their career, rather than graduates --

GADDY: I have an image of the group I feel responsible to -- my image may be more confusing than helpful.

VOYTUK: All you're going to present is facts -- you are not going to interpret them or make any inferences?

GADDY: What do you mean, facts?

VOYTUK: You are going to present data. How many people were employed or were seeking jobs or how many were going into industry or how many were going to into an academic positions?

GADDY: Yes. It's as factual as a survey is factual. I mean there's going to be some subjectivity in some of the questions. We haven't gotten as far yet as asking things like number of offers, and time to find jobs. These are fairly straightforward questions and may be factual, assuming they tell us the truth. I don't know if we will get into some things like perceptions of the job market, where it gets more subjective.

VOYTUK: Would you compare the data from 1998 with 1997?

GADDY: Yes.

VOYTUK: So, you are doing more than just presenting the facts; you will be doing some analyses.

GADDY: If the project lives -- which I hope it does -- that would be our goal. This would be the place to start. People who want more data can then go to the specific societies.

MITCHELL: I'm not really grasping it. How does this information help the person? What value does it have to someone who is about to enter the job market to know what happened to the prior year's graduates? Can someone say?

If you know, for instance, that 14 percent of last year's graduates went into jobs in industry, what are you going to do based on that information?

GADDY: Well, one of the places that this kind of information was recently requested was in the COSEPUP report. And I hear criticism of faculty for not warning their students about what they are getting themselves in for, because the faculty probably don't have the time to do the research. We do it full-time and can barely keep up. And I hear students saying, what is going on out there, would somebody please tell us?

We know what happened in 1993, but what happened last year? What is going on right now? So we are trying to bring it together, make it more consistent and get it out there a little bit faster to people.

I guess I would have liked to have known what happened to my predecessors -- and again, don't get too locked in on those three months from graduation? In their last two years of graduate school the data will only be one year old. You can start to get a sense of what's happening in your particular field in a small time window that you, as an individual trying to find a job, care about.

You don't care whether it is a 1.6 or 1.7 unemployment rate. You care about what has recently happened.

COSTELLO: I was going to say that I went to the University of Wisconsin where I received my Ph.D. in 1984. The entire academic culture there was based on the presumption that every graduate student was going to go to an equivalent university research institution. It never occurred to any of us until we were in the job market in that final dissertating year, that the expectation and the reality didn't match up at all. It would have been incredibly helpful to have a report that showed that for the previous six years only 7 percent Ph.D.'s in sociology from research institutions were going to major research institutions, , 25 percent were going to teaching colleges around the country. That gives you an idea, as a graduate student, that you are far more likely to have an option of teaching at a small college in the Midwest than going to the University of Michigan.

I think that kind of thing could be very helpful if, for no other reason, than to counter the prevalent cultures within all of the disciplines that presume only one outcome for their graduate students. That presumption and reality are quite different.

GADDY: It is also coming at a good time, too, when programs are starting to look more at outcomes either because accrediting bodies are asking them to or because their State legislatures are asking them, "What happened to your graduates?" And, Peter, I know you have got one project specifically looking at that. So there is more interest in looking beyond and seeing what happened. Some programs are incredibly good at it. Wasn't it the physics program at the University of Alaska that has a directory of photos of all of their graduates from the past 30 years, where they are, what they're doing, and their hobbies. It is lovely. Other schools maybe couldn't account for more than 5 or 10 percent of their graduates -- they don't even know what happened to them. So we are in a different era now which I think will be supportive of it and helpful too.

STREETER: I was glad to hear that you were trying to do that because I do get calls from all over the U.S. and foreign countries, too, about this. In fact, I got one about two weeks ago about this type of data. People not only want to know what happened to the students in the previous years, but what State they go to and the field they go into. That way they can direct themselves, maybe, to move out of State to find a job. I think that would be very helpful to the graduates.

KRUYTBOSCH: I thought that that was a very good observation that was made -- what kinds of institutions, what kinds employment places are people going out to? That would be very helpful to get your sense of where you should orient yourself. But I don't think that you would get that from the quick turn-around data. For example, those data did exist in 1984 -- the Carnegie surveys. I even gave a presentation called Down and Out From Harvard and Berkeley where the immutable law that you get a job at a lower prestige institution than you got your degree was well demonstrated. The data existed, but it wasn't disseminated and used, or widely known among people in sociology in higher education. But it seems to me for that kind of thing you need to have a richer data base to permit those analysis, because those patterns don't change radically from year to year.

GADDY: I guess what we would like to do is do this whole thing in a complementary way. We are certainly not going to recreate the National Science Foundation -- not on this project. But we'd like to be complementary and fill some of the holes. If it turns out that what we produce is positively identical to something that the NSF produces, then maybe it's redundant or maybe there are some elements that are redundant and we don't do it any more.

On the other hand, many of the societies now are doing survey efforts which tell me that there is something that someone is getting from it. And I would like to see if that is helpful if we could expand on it.

Also, from working with the society data, I know that there tend to be some field-specific observations that they can make that are way out of the scope of anything that NSF could ever have the time or the responsibility to worry about. My favorite example is that in psychology -- my home field before I got underemployed and became an executive director -- in the business sector we have the psycho-analyst in New York and the human factor psychologist doing cockpit design in Arizona in the same category.

Those are real different things to do for a living. If we can get some field-specific sensitivity like that I think it will help as well. This is impossible for NSF to do, given the scope of its surveys. We are going to try very, very hard to fill this gap and to do it on a consistent basis.

BAKER: Following up on what you said, I think what you are doing is very exciting and I think it fills an important need. I think there is a real challenge that all of us face in making sure that we are adding to the information and not getting caught in the middle of squabbles about, "Is it this percent or that percent?" I don't know how we are going to resolve that. I applaud SRS for taking the initiative, and the societies for doing these kinds of discussions. I hope that there are more of these meetings, because having been in the middle of some of those issues, as many of you have been, it is very tough to explain all the details about how this survey produced this number and that survey produced that number. To the outsider, that is useless. It drives them crazy, because they see us as these people in Washington. They think there are supposed to be national data whether it is a national association or Federal agency or something else. It drives them crazy and it drives policy groups crazy as, Ken well knows, to have different numbers or confusing numbers.

I don't know the solution but I wanted to make a plea that as we move forward on these multiple tracks that we, at least make sure that we are communicating . It is not likely that we are going to be able to find a way to make all the numbers match, but at least let's agree on who is going to say what on what. Because I would hate to see the discussion deteriorate into a battle of the numbers -- where we end up giving confusing information. I think these meetings are the perfect forum to have some of this conversation.