Beyond the dualism between lifelong employment and job insecurity : some new career promises for young scientists F. Dany EM-LYON 23 avenue Guy de Collonges BP 174 69132 Ecully Cedex Ph : 33 4 78 33 78 00 Fax : 33 4 78 33 79 27 E-Mail : email@example.com http://www.em-lyon.com
V. Mangematin INRA/SERD - UPMF BP 47X 38040 Grenoble Cedex 9 France Ph : 33 4 76 82 56 86 Fax : 33 4 76 82 54 55 E-Mail : firstname.lastname@example.org http://www.upmf-grenoble.fr/inra/serd
November 5, 2001
Abstract This article analyses the early careers of young scientists in France. Since training and early career management are designed to cater almost exclusively for an academic career, a substantial proportion of PhDs lack support to design their training in relation to the job they will look for after graduation. Even if most young scientists manage to find employment, this situation helps to reinforce the feeling that there is no future outside an academic career. At a time when the number of permanent posts is declining, it can penalise scientific activity by discouraging some from choosing this option. That is why this article will explore new directions for career management. Based on an analysis of the first ten years of 652 scientists’ careers in life sciences, it suggests that tenure is not the only possible promise for young academics. More precisely, it highlights the need for broader career promises, including career prospects outside the academic sphere. The introduction of such new career prospects would help to adapt research management to current realities of the scientific job market. It would also facilitate the diffusion of scientific knowledge in the economic sphere.
Key words : implicit contract ; academic fields ; PhD ; higher education ; incentive ; collaborative research ; university ; job market ; human resource management ; promise ; career.
1. Introduction The 50 years since the end of World War II have seen unprecedented growth in the life sciences. Recent years show a rapid increase in that process in France and in Europe, linked to the creation of biotech start-ups, the restructuring of the life sciences industry, and the priority in the 5th European framework programmes. The continued success of the life sciences research enterprise depends on the uninterrupted entry into the field of well-trained, skilled and motivated young people. Thus, for this critical flow to be guaranteed, it is necessary that young aspirants see that there are exiting challenges in life sciences research. Most of all, they need to have a reasonable expectation of having a good job after their long years of training. Yet recent trends in employment opportunities suggest that attractiveness to young people of careers in life sciences research is declining. Two major worldwide trends characterise this decline. The first one is the widespread hiring of scientists in soft money in academic institutions, i.e. positions that exist only as long as the scientists have funding from external sources that pays both the cost of running the lab and the cost of scientists’ salaries. The second one is the failure of many post-docs to find careers in the academic sector where they had assumed they would find long-term employment. These trends and the feeling of increasing precariousness of scientists’ careers could discourage good students from starting a PhD in life sciences and from working on senior scientists’ projects. In 1999, the National Research Council emphasised the increasing gap between the status of researchers with permanent posts and those with temporary positions. This makes the promise of an academic career less credible. According to this institute, the number and quality of candidates for PhDs are likely to decline considerably in coming years. In France, as PhD students represent 30% of the skilled manpower in life sciences academic teams (OST, 2000), a lower number of PhDs in life sciences would be critical : this decline may lead to a reduction of scientific production in life sciences as well as a slowing down of the diffusion of scientific knowledge (Mangematin, 2001). The annual report on doctoral studies in France (DGRT, 2000) reveals that a large proportion of life sciences PhDs do not have a permanent position two years after the end of the their PhD ; only 45% have found a permanent position in academia (29,4%) or in firms (15,5%). 55% of PhDs still have a short-term position, either as post-docs (46%) or unemployed. Such a situation is unusual in France for highly qualified workers. Therefore, the current trend concerning the careers of scientists raises two questions : • How to encourage students to do a PhD when the number of long-term jobs in academia is stable or decreasing ? • How to maintain essential collaboration between academics and PhDs if most PhD students have no future in academia ? Focusing on the issue of the management of scientists’ early careers, this paper explores the types of incentive which could be promoted to allow academic labs to continue producing quality research. It explores possible alternatives to the binary model on which research careers have almost exclusively been based until now, i.e. a life employment model as opposed to a market model and job insecurity. Following Stephan and Levin (1997) who call for a rethinking of scientific careers, it highlights new career promises (Dany, 2001) which may be used by academic labs to continue to attract good students to the PhD track. Building on Dany’s work on career promises, it explores the possibility to develop a new promise which is legitimate and credible, as well as legible (Dany, 2002) The first part presents the principles of career management in academia. It shows why academic careers have been considered as a preferred tool to monitor PhD behaviour and to enhance collaborative attitudes in the research programme. It also introduces the notion of "career promises" which could be used to enrich the "traditional idea of career" (Kanter, 1989 ; Arthur and Rousseau, 1996) (Kanter, 1989 ; which still prevails in academia but which concerns a smaller and smaller proportion of young scientists. Data collected on a population of 652 life sciences PhDs is presented in the second part. Information on young researchers’ early careers highlights the fact that despite their expressed preferences for an academic career, PhDs know very little about scientific careers. Their first experience will help them to discover progressively what they can expect and what they should do in order to succeed (Mangematin, 2000). This result, which challenges the idea of a career dependent on internal clues rather than on external ones (Alvarez, 2000), prompts us to assume that in a context unfavourable to permanent positions, new career promises can be used by academic labs to attract brilliant young scientists. In addition to the implicit contract mainly designed for PhDs who expect an academic position (Stephan and Levin, 1997), promise of employability (Dany, 1997 ; Gratton and Hope-Hailey, 1999) (Dany, 1997 ; Gratton and Hailey, 1999) outside academia appears as a possible incentive. We will show that this kind of promise is likely to be both "legitimate" and "credible". Discussion, in part three, shows that new career promises imply significant changes both in the dominant conceptual framework concerning careers and in early career management practices. At this stage scholars, like research teams, seem to be concerned only about the careers of those destined to work in academic research. But the development of promise of employability as an incentive mechanism can be effective only if this new promise has a high degree of "legibility". This argues for the definition of new methods of managing researchers and for new communication practices intended for them. 2. theory and hypotheses The literature highlights several characteristics of scientific activities which have led to the development of specific career management practices for academics. Both a need for security and independence, and the wish to recruit the best and to develop cooperation with young scientists, have been underlined to justify the development of an academic labour market with life-long employment for those who gain access to it. This present management is based on the assumption of the attractiveness of permanent posts in academia. It does not take into account other possible career promises which seem nevertheless more suited to the current context (3). 2.1. Specific needs for scientific activities Four characteristics of scientific activities lead to a specific organisation of academic careers. First, the randomness of progress in the field of knowledge requires a particular organisation of the academic labour market. Because the frontiers of knowledge are constantly being pushed back, no academic can keep all their knowledge in an entire discipline constantly up to date. As a result, they have to specialise, as shown by A. Siow (Siow, 1998). But at an individual level specialisation increases the risk of obsolescence of knowledge (McPherson and Winston, 1983). The need for incentives to specialise is all the more crucial in so far as academics may be averse to risk and uncertainties weigh on many of those who specialise on a particular subject. Because they are based on scientific gambols, some specialisations on the cutting edge of knowledge may be unfruitful : some subjects could be scientific deadlock after years of efforts. On the other hand, some research orientations could be more promising than expected. Second, because research takes time and because success cannot be guaranteed, researchers can feel themselves weakened by the success of others. Counter-productive behaviours may then appear and researchers could hesitate to recruit or to help brilliant young scientists. Thus, some scholars have claimed that to allow the academic system to renew itself, it is necessary to reassure researchers and not to increase competition amongst them (Carmichael, 1998). Tenure is a way to give academics the opportunity to hire brilliant young researchers without any fears for the position of researchers who already have a job. Third, the ability of academic teams to attract young researchers and to promote co-operation is particularly important. Scientific production is increasingly collective whereas scientists’ careers are individual. On the one hand, Stephan (Stephan, 1996) shows that the average number of authors of an article listed in the Science Citation Index rose from 2.52 in 1979 to 3.5 in 1993. She also notes that scientists who collaborate with others produce more and better quality work than those who produce alone (Stephan, 1996:1220). In this conception of collective research, the work of young researchers is crucial because it is complementary to that of their seniors. Their research, often largely empirically based, makes it possible to test a large number of hypotheses and thus accelerate the cumulative process. On the other hand, using the example of physics in the US, S. Gruner et al. (Gruner, Langer et al., 1995) show that the dynamics of scientific production in physics depends not only on the resources allocated to a few "star scientists" but also to the average age of researchers in the discipline and hence to recruitment opportunities. Fourth, scientific careers are also related to a cognitive division of labour. Based on a Poperian epistemology, T. Shinn (Shinn, 1988) reports on the specific contribution of young researchers. He emphasises the cognitive division of work by describing the different roles of juniors, seniors and "professors or directors" in the scientific community. While the juniors pay particular attention to anomalies likely to undermine an explanatory model, seniors work directly on the selection of models and the insertion of data in explanatory models. Professors focus on generalisation by working on fundamental and frequent phenomena. Finally, star scientists such as Nobel prize-winners or professors at prestigious universities are responsible for the production of models. Scientific production is thus seen as a whole, with the presence of different components required to maintain its dynamics. These four reasons plead for a specific career management in academia. 2.2. Management tools for scientists Despite some diversity, research around the world has been managed for a long time on a similar basis, roughly equivalent to that of the "tenure" system. France is no exception. French academics are hired after their PhD or post-doc in academic teams in universities or national labs. They can have a temporary position only for a limited period of time (2 years). After that, they are hired by universities or public sector research and tenured. They then have a civil servant status. Their salaries are determined by status (assistant professor or professor), are not negotiable and are the same in each university. Specific career rules have been set up to attract the best (despite the risks). Those rules are based on the assumption that a guarantee of independence linked to a guarantee of life-long employment would suit scientists and encourage them to do their best to be recruited by academics labs. This incentive, a priori less expensive than others (wages, for instance), is supposed to be attractive enough to stimulate a large flow of PhD students. It is also argued by scholars in favour of tenure, that the certainty of having a job for life encourages cooperation which may be discouraged by too much competitiveness (Laudel, 2001). The perspective of the promoters of tenure is finally consistent with that of scholars who defend internal labour markets. First, all of them see career as a useful tool to attract people (and to take into account their need for security, in particular). Second, they also see career as an efficient mechanism to control behaviour. Indeed, career shows people how they can succeed in a job and what they can expect for their success. It is like a mould that shows people what is possible and desirable for them to do (Thomas, 1983). In this view, guarantees offered to those considered as good researchers are crucial. They prove to everybody that specialisation is not so risky, at least for those protected by tenure. They also aim to show that cooperating with good researchers could be in the interests of one’s career. Analysing the "implicit contract" behind tenure, Stephan and Levin (Stephan and Levin, 1997) show how this way of managing careers contributes to research development. Tenure not only helps the supervisor to motivate students ; it has two other advantages. The first is based on the delayed nature of the reward. The PhD supervisor will have time to appreciate the real potential of the candidate and will finally help the young researcher to enter academia only if s/he considers that the PhD is good enough to be supported without damaging her/his reputation amongst colleagues. The second advantage of this system is that it is based on the reputation of the professor. "The reputational incentive of the professor not to cheat on the arrangement is straightforward. If the student is educated too poorly to be considered creditworthy as a future dean or provided poor information, the professor will cease to be able to attract top graduates and the source of labour, compensated well below its opportunity cost, will dry up." (Stephan, 1997 : 16). In other words, this work suggests that senior researchers too must continue to do their best despite tenure ; the alternative is marginalisation. Thus, even if the efficiency of life-long employment has been discussed (Osterman, 1984), some arguments still exist to defend the efficiency of tenure (as a form of academic career). Another advantage of tenure highlighted by (Carmichael, 1988) is the fact that imposing a deadline for the granting of tenure to academics avoids indefinite postponement of the particularly unpleasant task of dismissing young researchers whose performance failed to meet set criteria. This mechanism helps to reduce the effect of the asymmetry of information between researchers and the authorities in charge of the management of the academic labs. 2.3. Some hypotheses concerning the likelihood of new career promises Tenure, which loosely resembles a pyramidal system, works reasonably well as long as there is a growing demand for new academics. But, in a perspective of reduced numbers of available jobs in the academic community, the implicit contract is broken : a PhD supervisor could no longer reward the efforts of students who participated actively in collective scientific production. The disappearance of this reward is seen as all the more critical in that most often tenure is opposed to job insecurity. In this view, academic labs have no hope of involving students in PhD programmes and senior researchers can no longer attract the best students. While academics overlook all other possible forms of incentive regarding researchers’ careers, career prospects are nevertheless numerous (Dany, 1997). At least two main forms of promise can be identified which refer either to internal or external job markets. The first is consistent with the traditional promise of career, while the second corresponds to the concept of a promise of employability. Both are worthy of interest : different reasons could lead us to assume that each of these promises could suit young scientists. First, new career promises could be more interesting than traditional ones, for some people at least. In her previous work on the concept of a career promise, Dany (Dany, 1997) showed that individuals could prefer a promise of employability rather than a promise of career, i.e. of "a succession of related jobs arranged in a hierarchy of prestige through which persons move in an ordered and (more or less) predictable sequence" (Wilensky, 1961). Some people could be interested by "a change in the psychological contract whereby there is an accepted strategic exchange of loyalty and security for interesting work and opportunities to enhance portfolio careers with new skills and experiences" (Handy, 1990). This accepted strategic exchange could refer to new career expectations (Arthur, Inkson et al., 1999) consistent with the idea that boundaryless careers are better suited to changes in personal life (and in particular to the willingness to develop or use new skills, or to relocate, or find a new balance between private and professional life, for instance). This interest in a new conception of career could also be linked to the fact that career plans could be seen as inappropriate. Because of so many layoffs people have learnt to rely on themselves rather than on firms ; even those who have a job for life could have learnt to take advantage of new careers (versus bureaucratic careers) since inter-firm mobility could help them to enhance their career opportunities (see on this point Arthur and Rousseau, 1996 ; Arthur, Inkson et al., 1999). Some firms already mix internal career promises with other forms of career promise, to stimulate young graduates. This is the case in particular in accounting firms. Young professionals, for instance, are subjected to the "up or out" model (Hinings, 1993 ; Hinings, Greenwood et al., 1999) : either they have access to long-term employment through access to the partnership, or they have to leave. In this model which is supposed to allow standardisation of behaviours and intensity of work when results are not controllable, internal promise is an incentive only for a minority. In a similar perspective, tenure could be thought of as one incentive among others. But other incentives might be developed for some researchers. Furthermore, new career promises could be appreciated by people who do not really have specific career expectations. They could consider, from a pragmatic point of view, that this kind of promise is not so bad. This statement refers to the fact that individual capability to manage one’s career is still limited (Dany, 2002). Most often, people do not have enough information to shape their professional path on their own. Moreover, their willingness to succeed leads them to submit to management rules which could be seen as a form of protection (Abel, 1996). This submission is all the more common in so far as people do not systematically have stable preferences a priori (Festinger, 1957) (March, 1978) (Watson and Harris, 1999). Nevertheless, the strength of management rules depends on what Dany calls their legibility, credibility and acceptability, that is, in the case of career management rules : • Legibility, on the possibility for professionals to identify prevailing norms at a given time, • Credibility, on the effectiveness of these rules as regards management of career paths, • And legitimacy, on their advantages compared to competitive rules (i.e. compared to totally opportunistic behaviour, for example). To sum up, whereas researchers’ career management was previously thought of essentially in terms of the duality between security offered to some via tenure or else precariousness, alternative forms of promise (such as employability outside academia) could be envisaged. An analysis of the literature has shown that this position is consistent with the development of new career expectations. In this perspective, the legitimacy of career promises should meet the willingness of young scientists to become more actively involved in their careers. What they want is a "self-invented career" (Hall, 1976). But this proposition is also consistent with the fact that young scientists might have no career projects in minds. Career management practices will then tell them what they can expect, for they will ground their career expectations in the management practices observed all around them. In this view, the weight of management practices is crucial, much more so than the expectations young scientist might bring with them when they enter a research career. Therefore, in order to appreciate the legitimacy of a promise of a career outside academia, we suggest investigating young scientists’ career plans. More precisely, we will test whether : H1.1. Young scientists tend to become more and more actors in their careers and enter into research with career plans in mind or, H1.2. On the contrary, they appear to be "open-minded" to the career rules they will discover. Second, we will analyse the credibility of such a promise. This is all the more necessary in that, according to some scholars, a promise of employability is nothing but a fool’s promise (Gazier, 1989), that is to say, an empty promise (Rousseau, 1995). To be attractive, a promise needs to be consistent with career management practices. We will therefore study the main features of the professional situation in which researchers find themselves ten years after they enter a research activity. More precisely, we will test whether : H 2.1. Careers exist outside academia or H2.2. If, on the contrary, the safest career path is to progress into academia.
The promise of employability out of academia does not mean that young scientists will be recruited. Such a promise only supposes that careers are possible outside academia. This route does not go hand-in-hand with precariousness ; it does not concern only the less brilliant researchers. It could be thought of as an incentive because young scientists believe that it is both credible and interesting for them to expect to pursue their career outside academia. In the discussion, we will focus on the last condition necessary to implement a new promise : that of legibility. 3. A survey on life sciences PhDs About 10,000 PhD students in France have graduated in the life sciences during the past ten years (1988-1998). Universities deliver doctorates but PhDs do their doctoral research in an academic team which is specialised in a specific field. The survey focuses on PhDs who have been trained in the life sciences (plant and animal sub-fields). The team in which they work during their PhD belongs to a research institute called INRA (for agronomic research). About 1,800 PhDs have been trained in an INRA scientific team. A sample of 652 life science PhDs (most of whom specialised in plant and animal biology) has been studied to understand more broadly the management of PhD students and the possible tools to attract and manage the behaviours of young scientists. The survey analyses both the individual characteristics of the PhD student and the way they begin their career. 3.1. PhD student characteristics The survey on 652 life science PhDs who were trained in one of the INRA** research teams and graduated between 1988 and 1998, focused on the respondents’ career intentions at the beginning of their PhD, on the way they performed their research and valorised the results (publication, patents, training courses, innovations, etc.), and on the way in which teams took their aspirations into account. The survey was conducted in 1999 by postal questionnaire. The questionnaire was structured to reveal the way in which PhD students’ career plans, their subsequent jobs and their career management fitted together. It comprised seven parts : • Individual information (gender, age, initial training, etc.) ; • Background information on the PhD (nature and duration of funding ; PhD supported by a private partner - industry - or not ; type of academic team in which the student performed doctoral research ; relations with industry ; academic situation ; mode of valorising results ; teaching activities or not during PhD) ; • Information on modes of valorising PhD (publications, patents, etc.) ; • Information on career plans ; • Information on type and nature of first job immediately after graduation (temporary or permanent post, in the academic community or not) ; • Information on the type and nature of the job in December 1998, when the questionnaires were sent out (temporary or permanent post, in the academic community or not).
Table 1 presents the characteristics of PhDs in the sample. Table 1 : Characteristics of each population N=652 Life sciences Characteristics of PhD student % of women 50.2 % Average age at time of completion 29.5 years Graduated from grandes écoles or medical or veterinary departments 42 % Characteristics of the PhD research Average time taken to complete the PhD Mode % of PhDs completed after 4 years 3,5 years 3 years 90 % Grants or research assistance given to the PhD student by the Institute (INRA) Grants by the Ministry of Research Grants by industry Other grants 50 % 25 % 7 % 18% Collaboration with a private firm during the PhD research No collaboration with a private firm 22.4% 77.6% Distinction Highest distinction "with merit" Other 57 % 43 % Average number of publications during PhD research and before the first job after PhD completion 2,9 (min 0, max 16) Post-doc after the PhD 36% Professional trajectories Permanent position in academia 55,5% Permanent position in research in a private firm 9,7% Permanent position as professionals 12,4% Short term position and unemployment 22,3% Professional expectations at the beginning of the PhD To become an academic 70.7% To get a job in the private sector 9.2% No a priori choice 20.1%
Three important characteristics are thus revealed : 1)In life sciences, half of all PhDs are women. The average age at which the PhD is completed is 29.5. Two different kinds of educational trajectory are drawn. The first describes students who graduated from grandes écoles or medical or veterinary departments i.e. institutions which are allowed to select the best students at entry to university. By contrast, public universities are not allowed to select students at entry. Selection is therefore weaker and is done step by step. On the whole, training at the grandes écoles is more intense and effective than that at university. Students from grandes écoles therefore tend to have more precise career plans than those from universities. 2)1) PhDs are hired by the scientific team to do their research. PhDs done in the Institute are generally of a better quality (large proportion of highest distinction and average number of publications) than the average in life sciences PhDs (DGRT, 2000). Other characteristics (average age at graduation, duration of PhD, etc.) are similar to those of all life sciences PhDs. 3)2) Over 70% of PhDssay they would like to have an academic position after graduating. Yet only 55% of them actually do so, in France or abroad. 3.2. Characteristics of labs and management of their young scientists The diversity of teams and the heterogeneity of each research team appear as one of the main characteristics reflecting a sharply contrasting image of scientific production. Recognising the diversity of research activities and defining quantifiable indicators for measuring results are the first steps towards the identification of research teams as a constituted entity. The characterisation of their production is a necessary although insufficient condition for focusing analysis on research teams. As shown by studies on the production of research teams, by Laredo et al. (Laredo, Mustar et al., 1992) and Joly and Mangematin (Joly and Mangematin, 1996), these aspects are characterised by teams’ involvement in each of the following four dimensions. The first is the most well known, i.e. involvement in the academic community. Academics are professionals who aim to produce scientific knowledge and to diffuse it. But researchers also participate in a process of creation of economic value resulting in innovation, in activities of general interest whose aim is more diffuse, and in activities related to popularisation and expertise. The academic dimension is usually identified by the team’s number of academic publications and the participation of each of its members in academic life (membership of scientific journal editorial committees, evaluation committees, etc.). Academics also participate in technology transfer and they are involved in innovation with industrial partners. Involvement in industry is identified by research contracts between an academic team and a firm, consultancy, expert missions by members of the team to firms, and PhD students funded by firms. The next two dimensions are important whereas they are more difficult to characterise. Academics are involved in science management and scientific leadership on behalf of public authorities. This is characterised by the number of research grants that the teams have with regional, national or European authorities, by participation in the selection committees of projects financed by public authorities, and by expert missions to ministries. Finally, academics play a key role in the training of students. Their involvement in education is characterised by the number of courses given by members of the team and the diversity of training in which they are involved. However, in our study, this dimension is not really relevant because most of the researchers are full-time researchers and do not have teaching duties.
Table 2 : Perceived characteristics of labs in which the PhD research has been done Involvement in academia Total Weak Strong Involvement in Weak 257 (44%) 195 (33%) 452 (77%) Industry Strong 56 (10%) 79 (13%) 135 (23%) Total 313 (53%) 274 (47%) 587 (100%) Missing value : 37, Chi square : 9.875, prob : 0.002 Table 2 reflects PhD students’ perception of the involvement of their lab. It is an individual perception which varies from one student to another for the same lab, depending on with whom the PhD research has been done. However, for almost all labs, students have the same ranking. When radical changes occur, such as turnover of permanent researchers or changes in research themes, the perception may change from one PhD student to another. When one PhD student and one lab are associated, it is possible to define four types of lab based on a combination of the two relevant dimensions : • Closed labs : 44% of academic teams appear to have a weak involvement both in academia and in industry. These labs are said to be closed because they have a weak involvement in any network. • Academics labs : The second type of lab is those which are mainly academic (33%). They are not really involved with industrial relations but have a high visibility in academia. • Technological labs : A small proportion of labs develops strong links with industry (10%). Their core competencies are mainly technological rather than scientific. • Open labs - as opposed to closed labs. The 13% of open labs are involved both in private partnerships and in the academic community. The type of lab is supposed to influence the management of young career practices because different types offer different profiles of relationship and training to their PhD students.
4. Results Among the main results which may support the idea that new career promises should be used to attract young graduates to a PhD, we can highlight first those which reinforce the view that some young scientists have no well-defined career plan. In fact, these are individuals who are prepared to redefine their career in relation to the opportunities offered by the laboratories described in our study. Their aspirations when they are recruited by research laboratories are not as strong as real career plans. These young scientists fail to take into account the realities governing scientific careers and do not seem to be steering their career when they start their PhD. They mostly accept positions offered to them and their main priority seems to be geographic location. 4.1. Seeds of reflection on career plans Even if almost 80% of PhD students expressed career preferences (see above, Table 3), the survey reveals that this is more a matter of emergent ambition than career plans. Indeed, two main elements attest to a lack of actual career plans or strategies : (1) criteria on which the choice of academic team is based and (2) the discrepancy between the PhDs’ stated professional plans and the competencies developed in the teams in which they do their doctorates. In fact, the survey reveals that students have little knowledge of how the world in which they would like to work operates. Table 3 : PhDs’ expectations toward career Wish to find a position in Industry Academic research No preference Total Total 54 418 119 591 Percentage 9.1 70.7 20.1 100% Missing values : 61 4.1.1. Criteria on which choice of an academic team is based Whereas scientific excellence is the most important criterion in academia, criteria of scientific visibility governed the choice of research team for doing their PhD for less than a quarter of the PhDs. Moreover, Table 4 reveals that most often PhD students do not choose their team but are chosen by the PhD supervisor. When PhDs mention criteria used to choose a team, they refer less to labs’ performance than to geographic proximity. Table 4 : Criteria for choosing team for PhD Not important Important Scientific visibility 78% 22% Geographic proximity (same university as for masters’ degree, practical work, etc.) 73% 27% Chosen by the supervisory team 48% 52% Criteria related to their employment (PhD funded by a firm related to the team) 85% 15% N=591, missing values = 61 Two complementary explanations can be suggested to account for this poor strategic behaviour of PhDs when they choose the academic team for their PhD. First, it is difficult for PhDs to choose their team. Labs have money to hire PhDs but the market is not really open and it is difficult for both PhDs and team leaders to hire PhDs from another university. Second, PhDs have poor information about the scientific competencies of teams before the beginning of their PhD. Tables 5 and 6 cover this point.
In fact, recruitment of PhDs resembles a process of co-option in which teams choose students rather than being chosen. In this process, the beginning of PhD students’ careers seems to be unrelated to their career plans and primarily guided by their performance (ranking) in the school system. It is only during their doctoral research that PhD students discover prevailing evaluation criteria in the academic community. Thus, although they do not seem to control their choice of team entirely, the criteria for doing post-doctoral research change substantially, as shown in Table 5. Table 5 : Evolution of criteria of choice between PhD and post-doc Amongst those who consider the following criteria as important to choose their post-doc team Considered as important to choose the PhD team Considered as non important Unknown Total Scientific visibility 35% 55% 10% N=105 Geographic proximity (same university as for masters’ degree, practical work, etc.) 30% 61% 9% N=89 Total number of PhDs who have a post-doc position and who answer to this question : N : 166
Progressively, scientific visibility becomes an important criterion in the choice of a post-doc team. 55% of those who did not consider this criterion as important in choosing their PhD team took it into account when choosing their post-doc team. Geographic proximity criteria correspond to an evolution in the way of living. Around 30 years old, PhDs are at an age where they start to have children ; it is therefore more and more difficult for them to move. 4.1.2. Discrepancies between the type of competencies developed by teams and PhD students’ career plans Research teams’ modalities of recruitment (selection of best students) and criteria used by students to choose their PhD team (criteria of proximity) generate discrepancies between the type of competencies developed by teams and PhD students’ career plans. The competencies developed by teams are evaluated in terms of their involvement in both the academic community and industry (the variables of which are defined precisely below). Table 6 compares PhD students’ career plans and the type of competencies (involvement) that the teams develop. Table 6 : Correspondence between the team’s involvement and PhDs’ PhDs’ expectations toward career Career plans in industry No a priori choice Career plans in academia Total Academic lab 9 (-8) 32 (-5) 145 (13) 186 Open lab 14 (7) 11(-4) 49 (-4) 74 Closed lab 17 (-6) 58 (9) 173 (-4) 248 Technological lab 11 (6) 9 (-1) 31 (-5) 51 Total 51 110 398 559 N:559, DF : 6, Value 27,2 Prob : 0,001 Despite a global link between the characteristics of the lab and the PhDs’ expectations toward career, Table 6 shows that the number of students in a closed lab (44%) remains high. Those who would like to make a career in industry or in academia can find themselves in laboratories whose characteristics do not correspond to their expressed preferences. This situation might reveal a lack of information for PhDs to choose their team. It may also reflect an absence of choices for PhDs. Either way, the fact that only 23 of the 51 who expressed their wish to work in industry do their PhD in a team with strong involvement in industry, and that close to 50% of those who intend having an academic career do their PhD in a team with weak involvement in the academic community, suggests that the expressed preferences of young scientists do not determine the choice of a lab, i.e. the choice of the kind of on-the-job training they will have and the kind of networks to which they will be connected. Young scientists are not reluctant to work with labs whose competencies do not fit their expressed career preferences. In other words, young scientists appear not to make strategic choices concerning the management of their career, at the beginning at least. They do their PhD in the lab which agrees to recruit them without paying much attention to its competencies. They seem to postpone their reflection on career plans and strategy. H1.2 rather than H1.1 tends to be confirmed. Indeed, changes in this attitude occur only when they become more aware of the rules governing researchers’ careers. Doctoral studies appear to be a time of discovery of the scientific world. But observed behaviours suggest that preferences initially expressed for "academic" careers possibly reflect a dominant norm as much as a real personal goal. These preferences are expressed by young people who have only vague knowledge of the realities of scientific activities ; above all, their capability to manage their career seems very limited. To conclude on this point, the expressed preferences concerning career look more like emergent ambitions than actual career plans. The results supported the idea that young scientists remain open to different careers at the beginning of their PhD. Thus, it seems that young scientists rely on management practices to orient their career, and that they progressively become actors of their career, as they obtain more information on what is possible for them to do. In this view, a promise of employability could replace a promise of a career in academia. This type of promise could be emphasized for those who want to go into research but do not know how to manage this kind of career. A promise of employability could also attract those who, pragmatically, grasp the opportunities offered to them, even if it leads them to join less well-known labs with weak ties with the networks likely to interest them. This implies, however, that this type of promise is credible and that students believe in it and are prepared to comply with prevailing norms because they are not afraid of job instability. 4.2. The credibility of the promise of employability Assuming that PhDs, like many others, tend to avoid job insecurity, we suggest first analysing the determinant of precarious positions. This will enable us to show that, contrary to what one might imagine, a career promise outside academia is not less credible than one in academia. This is in any case what emerges from the analysis below on the most sure individual plans (1) and on the conditions of access to jobs (2). 4.2.1. The determinants of precarious positions The mean time in which PhDs occupy a temporary position after their degree is 7.2 months with high disparities (Std Dev : 8.15). Table 7 reveals that 50% of PhDs have to face precarious positions for more than 6 months after their PhD. A transition time is usually required after the PhD to find a job. As the end of the PhD is usually a rushed time (dissertation defence, publications, etc.), we consider that six months is a normal interval between the end of the PhD and a new job. However, when compared with other degrees such as engineering degrees, the employment situation of life sciences PhDs appears to be far less stable. This is mainly due to the modes of recruitment in academia. Table 7 : Distribution of time in temporary jobs (or unemployment) after the PhD Number of PhD Percentage Cumulative percentage 0 months 240 38.5 38.5 <6 months 71 11.4 49.9 6<<12 months 193 30.9 79.8 > 12 months 120 19.2 100.0 N : 624 ;
To understand and evaluate the relative role of each variable on the permanent or temporary position, we use a multinomial logit model of determining factors. The logit model analyses the influence of different qualitative variables on qualitative variables to be explained (Maddala, 1983). An endogenous variable is "Temporary position for more than 6 months at the end of the PhD versus permanent position". Tested before using a chi square analysis, two variables seem not to have any influence (career plans and type of labs). They have been withdrawn. Table 8 : Influence of each variable on the permanent or temporary position Current employment Coefficient Std error Probability Significance Constant Temporary > 6 months -0.05 0.10 0.26 0.61 Permanent jobs Publications >2 Temporary > 6 months 0.14 0.08 2.72 0.10 2 Permanent jobs Gender Fem Temporary > 6 months 0.10 0.08 1.45 0.23 Masc Permanent jobs Funding Firms Temporary > 6 months -0.43 0.16 6.72 0.00 Grant Temporary > 6 months 0.27 0.12 4.71 0.03 INRA Permanent jobs Post doc Yes Temporary > 6 months 0.73 0.09 63.4 0.00 Post doc No Permanent jobs Ratio of the max. of likelihood DF : 18 Chi square : 13.4 Prob : 0.77
Table 8 shows that three variables influence the fact of having a permanent or a temporary position : • Having a grant or a salary from industry during the PhD reduces the risk of a temporary position • A post-doc position increases the risk of having a temporary position after the post-doc • Having a grant from university increases the risk of a temporary position. We can add that writing more than 2 publications is not a guarantee for having a permanent position. Women seem to be subjected to negative discrimination. More women than men have temporary jobs. This requires two comments : • Those PhDs who set their sights on an academic career are the ones who are most likely to find themselves with temporary jobs. This result is valid even for those whose personal competencies seem good and who publish as much as the others. It seems that this situation has two main explanations. First, recruitment in the academic sector is spread over a long period, from February to September, and PhDs often have a temporary job while waiting for a permanent one. Second, difficulties of finding jobs for PhDs in the life sciences (at least over the period 1990-2000) relates to a downturn in recruitment in the academic sector and little recruitment in the industrial sector. During a period in which positions offered to those aiming for academic careers are rare, candidates’ qualities are not enough to guarantee them permanent jobs. Under these conditions, academic careers seem less predictable than other types of career paths oriented towards the private sector. Contrary to prevailing representations, the academic career is not necessary a guarantee against job instability. This invalidates H2.2. • Despite the suspicion it may arouse, a promise of employability is not necessarily a fool’s promise. While academic careers have been developed to fit scientists’ presumed need for job security, our survey reveals that the safest way to get a permanent job in research is to agree to work outside academia. Only 26% of those who intend to develop their career in industry experienced unemployment or are forced to take temporary jobs. This is the lowest rate. Despite no significant linkages between career plans and temporary position, we can add that working in a lab with a strong involvement in industry (open or technical) is the best way to avoid a temporary job or unemployment. These labs offer more security than those with strong involvement in the academic community. 4.2.2. Determinants of job when interviewed Table 9 specifies the above elements. It shows that the characteristics of the lab in which the PhD is done, the career ambitions of the PhD, the fact of having a post-doc or not, and collaboration with industry during the PhD, all play a decisive part in the PhD’s chances of obtaining a particular job. The logit model analyses the influence of different qualitative variables on qualitative variables to be explained (Maddala, 1983). The endogenous variable is "current job at the time of interview". The exogenous variables are listed in the table : characteristics of the lab in which the PhD is done ; career ambitions of the PhD ; number of publications ; gender ; collaboration with industry during the PhD ; and the fact of having a post-doc or not.
Table 9 : Analysis of variables that influence current employment Current employment Coefficient Standard error Probability Significance Constant Other perm.jobs -1.46 0.22 0.00 *** Temporary -0.47 0.15 0.00 *** Private research -1.94 0.24 0.00 *** Academic research Type of labs Academic Other perm.jobs -0.64 0.25 0.01 *** Academic Temporary -0.06 0.17 0.75 Academic Private research -1.19 0.42 0.01 *** Open Other perm.jobs 0.04 0.28 0.59 Open Temporary -0.14 0.24 0.39 Open Private research 0.86 0.32 0.00 *** Technological Other perm.jobs -0.00 0.32 0.98 Technological Temporary -0.14 0.27 0.59 Technological Private research 0.56 0.37 0.13 Close Academic research Publications 2 Other perm.jobs -0.21 0.18 0.26 2 Temporary -0.08 0.14 0.59 2 Private research -0.15 0.27 0.58 3-5 Other perm.jobs -0.17 0.19 0.38 3-5 Temporary -0.15 0.15 0.29 3-5 Private research -0.04 0.28 0.89 +5 Academic research Gender Fem Other perm.jobs -0.04 0.13 0.74 Fem Temporary 0.23 0.10 0.01 *** Fem Private research -0.00 0.19 0.98 Masc Academic research Post doc No Other perm.jobs 0.84 0.18 0.00 *** Post doc No Temporary -0.12 0.10 0.23 Post doc No Private research 0.28 0.21 0.19 Post doc Yes Academic research Collaboration No Other perm.jobs -0.48 0.15 0.00 *** No Temporary -0.23 0.12 0.05 ** No Private research -0.50 0.20 0.01 *** Yes Academic research Ratio of the max. of Likelihood DF : 444 Chi-square:375 Prob 0.99
Table 9 shows that : 1) laboratories with little or no involvement in research partnerships with the socio-economic world produce students who will have difficulty finding employment outside the academic world ; in other words, being a PhD in an academic lab decreases the probability of finding a job in industry ; 2) Thus, the laboratory’s research competencies influence the PhD’s future. That said, we note that the competencies that laboratories develop through collaboration with industry is not enough to offer PhDs a significantly higher probability of finding a job in the private sector. Industry wants researchers with recognised competence in the academic world. Thus, the most sought-after candidates in industry are those from open laboratories who benefit from the relations that their laboratory maintains with industry and who also have a recognised academic reputation. This importance of academic recognition explains why PhDs from open labs do not have fewer chances than others of obtaining a job in academia. Academic excellence affords access to jobs ; the nature of the lab’s relations allows openness towards the industrial world. These results enable us to complete available information on the credibility of a promise of employability by highlighting a number of criteria determining the effective employability of PhDs. This, in the final analysis, depends on : 1) The academic quality of their work. Contrary to what one might expect, those who aim for private sector research are not necessarily the less brilliant researchers. Their number of publications is equivalent to that of PhDs who remain in academia. 2) Relations they have personally developed with a view to making a career outside the academic world. The PhD is a period of on-the-job training during which PhD students acquire know-how and skills similar to those of the people training them. Thus, the competencies of members of a laboratory are diffused. At an individual level, however, the PhD can decide to nurture a particular type of relationship. For example, having maintained cooperation with an industrial partner during the PhD increases a PhD’s chances of finding a job in the private sector. The PhD thus appears as a time of construction of a sphere of competencies and relations that influence the graduate’s employment at the end of their PhD (Mangematin, 2000). 3) The lab’s relations with the private sector. By contrast, doing a PhD in a team that is heavily involved in academic production does not guarantee a job in academic research. Here, the "breeding ground" effect comes fully into play.
To conclude, "putting the stress on a promise of employability" does not mean that young scientists will be taken in. Our survey simply allows us to affirm that : "some careers are possible outside academia", at least for those who have suitable experience. In other words : "the development of such a promise does not necessarily go hand-in-hand with more precariousness". At least two kinds of promise could be promoted : one for those more interested in academic careers and another for those who want to do research but who may be interested by other activities than academic research. Empirical findings show that, not only do careers exist outside academic research (validation of H 2.1) but that PhD’s career plans evolve substantially during their PhD (validation of H 1.2). Thus, research management practices could have a strong influence on this population which is concerned about its professional future. 5. Conclusion The results presented above painted a picture of PhD students who lack sufficient in-depth knowledge of the operational modalities of research activity. Contrary to the assumption made by Stephan and Levin (1997, see above, part 1), they do not master their career. Moreover, even though they are able to express preferences concerning their future, those preferences do not prevent them from joining labs whose competencies do not suit their presumed individual projects. Their preferences for an academic career are not as strong as they claim. Thus, it seems possible to use a promise of employability to get them involved in academic research despite the declining number of permanent jobs in academia. The interest of such a promise does not only refer to the fact that different career paths could suit students. It is also supported by the fact that this kind of promise is more consistent with the current reality of the labour market for researchers. Contrary to what is often assumed, being engaged in research with industry could be safe, and even safer, than being strongly involved in the academic community. These results, which do not deny the interest of a promise of an academic career, nevertheless have some strong implications both for future research and for career management in academic labs. 5.1. Implications for future research First, the fact that students could embark on a research career with little knowledge about the career rules prevailing in this field makes it necessary to know more about the very nature of their motivation. The aim of such research would be both to avoid discrepancies between their motivations and the nature of a research career, and to know more about the different career anchors (Schein, 1975) on which promises concerning the career could be based. In particular, some meta-analysis could be done in order to know more about the issues which warrant consideration, in addition to that of "how to gain access to a permanent job" which has been dominant in the literature. Perhaps the attractiveness of research activity could be based on other points. Second, while scholars could see a promise of employability as dupery (Gazier, 1989), our results prompt us not to overlook this kind of promise too quickly. The data show that some teams can be seen as very good access to employment in the private sector. The management of a promise of employability, like the management of in-house careers, could help people to gain advantages which make their futures relatively enviable. Although without tenure (or titularisation for the French case) people cannot have a watertight guarantee of a career, training as well as rich experience or interesting networks (Lagarenne and Marchal, 1995 ; Eymard-Duvernay and Marchal, 1997) enable them to accumulate advantages in the various "career tournaments" (Rosenbaum, 1990). While senior academic researchers might disregard careers in industry, and while the dominant post-modern literature on career puts the stress on career management by individuals, it is important to remember that organisational practices concerning career could help to enhance people’s employability. In this view, research dealing with the management of employability could be very useful to make this promise more credible in the academic community and to encourage some new practices in academic labs. 5.2. Implications for career management in academia. Whereas job insecurity is expected to increase, labs still do not pay enough attention to the careers of the young scientists who join them to do their PhD. As Table 8 shows, only a minority discusses their career plans with members of their team at the beginning of the PhD. Table 8 : Career plans and help from the team No interaction with a member of the lab about career plans Interaction with a member of the lab about career plans Total PhD student’s career plans Industry Academic research No preference Total Industry Academic research No preference Total Discussion of career plan with a member of the lab at beginning of PhD 31 5% 176 30% 94 16% 301 51% 23 4% 242 41% 25 4% 290 49% 591
Discussion of career plan during the PhD 36 6% 221 37% 83 14% 340 58% 18 3% 197 34% 36 6% 251 43% 591
Help of a member of the lab to define career plan in more detail 44 7% 263 45% 78 13% 385 65% 10 2% 155 26% 41 7% 206 35% 591
N=591. Each Chi square by line are significant This lackof interactionaroundcareerplansis particularly noteworthy in the cases of those PhDs who have no idea of what they want to do after their thesis and those who intend to work in industry. In other words, the only cases that seem to interest the laboratories are PhDs who might join the academic community and become future partners or colleagues. Yet, even if this runs counter to prevailing representations in the academic world, the results of the study on the credibility of the promise of employability show that several alternatives exist after the PhD, in both academia and the private sector. The promise of employability in the private sector is therefore credible because it already functions for over 20% of PhDs. It is the PhDs who think about their futures and those who want to go into the private sector who are likely to be the most interested in these promises of employability, and for whom laboratories could create appropriate early-career management. The main problem concerning the promise of employability is its legibility. This is reduced because, even today, career management discourse and practices limit prospects to academia. They give the impression that outside academia there are no alternatives to job insecurity. Yet simple tools such as a directory of PhD graduates could illustrate the different possible careers, and thus put an end to the caricatural dualism between lifelong employment (tenure) and job insecurity. To conclude, the changes we are calling for both in the conception of researchers’ careers and in career management practices aim to improve attractiveness of an investment in a PhD. Apart from the fact that such changes could help to challenge young scientists’ feeling of increasing insecurity, we can underline another benefit. New careers could also be seen as a good way to improve transfers between the academic and industrial worlds. The flow of knowledge between organisations is based on the circulation of people and PhD students in particular since they are the most mobile part of the scientific population. Circulation of young scientists between organisations and between research teams enables host organisations to acquire knowledge produced in other teams. In this case, acquisition of knowledge and know-how is achieved not through replication of experiments carried out elsewhere but through the temporary or permanent employment of a researcher. Thus, collaboration between academic teams and firms generates an exchange of PhD students or technicians. When firms or teams recruit qualified staff with specific competencies lacking in the organisation, this recruitment corresponds to the diffusion of knowledge from the university or team that trained the individuals. Movements between academic and private research are still exceptions in France. Diffusion of knowledge through the exchange or recruitment of staff takes place very early, just after the PhD or post-doc, or not at all. Training young researchers able to find jobs in the private sector is an important challenge for French academic research.
Bibliography Abel, O. (1996). Paul Ricoeur. La promesse et la règle,. Paris, Michalon. Alvarez, J. L. (2000). Theories of Managerial Action and their Impact on the Conceptualisation of Executive Careers. Career Frontiers. M. Peiperl. London, Oxford University Press : 127-137. Arthur, M. B., K. Inkson, et al. (1999). The New Careers. London, Sage. Arthur, M. B. and D. M. Rousseau, Eds. (1996). The Boundaryless Career, a New Employment Principle for a New Organizational Era. New York, Oxford University Press. Carmichael, L. H. (1988). "Incentives in Academia : Why is There Tenure ?" Journal of Political Economy 96(3) : 453-472. Dany, F. (1997). La promesse d’employabilité : un substitut possible à la promesse de carrière ? Département des sciences de gestion. Lyon, Université Lyon III : 514. Dany, F. (2001). Beyond Psychological Contracts : the Concept of "Career Promise". A useful concept to analyse the current careers ? 17th Egos Colloquium, Lyon (FRA), EM Lyon. Dany, F. (2002). "Cadre "sujet" ou cadre "contraint" : le cadre contemporain face aux pratiques de gestion des carrières des grandes entreprises." Revue Française de Gestion Forthcoming. DGRT (2000). Rapport sur les études doctorales 1999. Paris, Ministère de l’Education nationale, de l’enseignement supérieur et de la recherche. Eymard-Duvernay, F. and E. Marchal (1997). Façons de recruter. Paris, Métailler. Festinger, L. (1957). A Theory of Cognitive Dissonance. New York, Harper & Row Publishers. Gazier, B. (1989). Employabilité de crise et crise de l’employabilité. Séminaire d’Economie du Travail, CNRS-URA n°919. Gratton, L. and V. Hope-Hailey (1999). The Rhetoric and Reality of New Careers. Strategic Human Resource Management. L. Graton, V. Hope-Hailey, P. Stiles and C. Truss. New York, Oxford University Press : 79-100. Gruner, S., J. S. Langer, et al. (1995). "What future will we choose for physics ?" Physics today December : 25-30. Hall, D. T. (1976). Careers in Organizations. Glenviewi11, Scott, Foresman. Handy, C. (1990). The Age of Unreason. Harvard, Harvard Business School Press. Hinings, C. R. (1993). The emergence and transformation of organisational Archetypes : institutional, task and economic dynamics. 11th EGOS Colloquium, Paris. Hinings, C. R., R. Greenwood, et al. (1999). The dynamics change in large accounting firms. Restructuring the Professional Organization. D. Brock, M. Powell and C. R. Hinings. London and New York, Routledge : 130-153. Joly, P. B. and V. Mangematin (1996). "Profile of public laboratories, industrial partnerships and organization of R and D : the dynamics of industrial relationships in a large research organization." Research Policy 25(6) : 901-922. Kanter, R. M. (1989). When Giants Learn to Dance. New York, Simon & Schuster. Lagarenne, C. and E. Marchal (1995). "Recrutement et recherche d’emploi." La lettre d’information du Centre d’Etudes de l’Emploi 38. Laredo, P., P. Mustar, et al. (1992). Defining the strategic profile of research labs : the research compass card method. Science and technology in a policy context. A. F. J. Van Raan. Leiden, DSWO Press. Laudel, G. (2001). "Collaboration, Creativity and Rewards : Why and How Scientists Collaborate." International Journal of Technology Management 22(8). Maddala, G. S. (1983). Limited dependent and qualitative variables in econometrics. Cambridge, MA, Cambridge University press. Mangematin, V. (2000). "PhD Job Market : Professional Trajectories and Incentives During the PhD." Research Policy 29(6) : 741-756. Mangematin, V. (2001). Les docteurs, producteurs et diffuseurs de connaissances. Encyclopédie de l’innovation. P. Mustar and H. Penan. Paris, Economica. à paraître. March, J. (1978). "Bounded Rationality, Ambiguity and the Engineering of Choice." Bell Journal of Economics 9(2). OST (2000). Science & Technologie : indicateurs 2000. Paris, Economica. Osterman, P., Ed. (1984). Internal Labor Markets. Cambridge (Mass), MIT Press. Rosenbaum, J. E. (1990). Organisation career systems and employee misperception. Handbook of career theory. Cambridge, Cambridge University Press : 329-353. Rousseau, D. M. (1995). Psychological contracts in organizations : Understanding written and unwritten agreements. Thousands Oaks, CA, Sage. Schein, E. H. (1975). "How Career Anchors’ Hold Executives to their Career Paths." ANDCP-Personnel 3 : 11-24. Shinn, T. (1988). "Hiérarchies des chercheurs et formes de recherche." Actes de la recherche en sciences sociales 74 : 2-22. Siow, A. (1998). "Tenure and Other Unusual Personnel practices in Academia." Journal of Law, Economics and Organization 14(1) : 152-173. Stephan, P. (1996). "The Economics of Science." Journal of Economic Literature 34 : 1199-1235. Stephan, P. E. and S. G. Levin (1997). "The critical importance of careers in collaborative scientific research." Revue d’economie industrielle 79(1) : 45-61. Thomas, B. A. (1983). "Managerial careers and the problem of control." Social Science Information 22(1) : 1-25. Watson, T. and P. Harris (1999). The Emergent Manager. London, Sage. Wilensky, H. L. (1961). "Careers, Lifestyles, and Social Integration." International Social Science Journal 12 : 553-558.