MONETARY REWARDS AND FACULTY BEHAVIOUR - CiteSeerX

6 downloads 0 Views 38KB Size Report
HOW ECONOMIC INCENTIVES DRIVE PUBLISH OR PERISH ... services will in the future be crowded out by 'publish or perish' which itself is induced by ...
SMA_2008_Proceedings-Page0725

MONETARY REWARDS AND FACULTY BEHAVIOUR: HOW ECONOMIC INCENTIVES DRIVE PUBLISH OR PERISH Uschi Backes-Gellner, University of Zurich ([email protected]) Axel Schlinghoff, University of Bonn ([email protected]) ABSTRACT Higher education institutions are increasingly emulating research institutions and concentrate their financial rewards on research output (Melguizo/Strober 2007). We investigate whether faculty members strongly react to such financial rewards. We focus on monetary rewards originating from promotions. Based on economic theories we derive three hypotheses. We test them with data from 112 faculty member from the USA and 189 from Germany. Consistent with our hypotheses we find that faculty members strongly behave like economic theory suggests: they not only increase publications when monetary incentives are large but they also instantly cut them down as soon as monetary incentives are gone. Introduction As Melguizo and Strober (2007) show faculty members are financially rewarded for enhancing institutional prestige, consistent with what emerging economic theories of higher education would predict. One consequence according to economic theory is that faculty members will spend more time on prestige enhancing research outputs. Since spending more time on teaching (or other activities) has no effect on salary even in comprehensive universities or liberal arts colleges, faculty members, according to the same economic reasoning, can no longer be expected to exert high efforts on these other activities including teaching, advising or community services. This in turn means that teaching or community services will in the future be crowded out by ‘publish or perish’ which itself is induced by universities’ ever increasing attempts to raise institutional prestige. So if faculty members really behave as expected and react to economic incentives as assumed, current reward systems in higher education cause serious and growing problems because as Melguizo and Strober (2007) also point out, all types of higher education institutions are more and more emulating research institutions in their pursuit of prestige meaning that monetary rewards for prestige maximizing research activities will become ever stronger. Accordingly, faculty members are induced to slight activities such as teaching, advising and community services causing their institutions to drift away from important parts of their higher education mission. In our paper we investigate whether faculty members do indeed behave according to what economic theory suggests and whether they significantly adjust their behaviour when monetary rewards are changed. We focus on promotion induced monetary incentives, i.e. pay raises attached to a promotion for example from assistant to associate or full professor. We study over the lifetime of individual faculty members whether their output reacts to changes in individual incentives. We expect research output to be higher in a situation where monetary incentives to produce research output are high (vice versa we expect research output to drop as soon as monetary incentives become smaller). If on the other side faculty members were only intrinsically motivated we would not expect this type of instant and monetary related output changes. So systematic output changes before or after a promotion event give a clear indicator on faculty reaction to economic incentives. The advantage of our empirical method, which compares intra-individual productivity differences and monetary incentives over a single researcher’s career, is that productivity results are not biased by inter-individual ability differences as would be the case in a cross-sectional comparison of differences

SMA_2008_Proceedings-Page0726

between individuals and their respective research output and monetary incentives. By studying individual output changes we are able to detect whether career incentives drive the well-known ‘publish or perish’ phenomenon. Theoretical Analysis As is known from a large number of studies in personnel economics the strongest financial rewards in most organizations arise from promotion induced wage increases and not from piece-rate-like pay or bonus pay. This general pattern is also found in universities where pay increases with tenure and other promotion steps are also among the most important determinants of individual income variation. To study the effect of such career related monetary rewards economists use tournament theory because it provides a number of fruitful predictions. Tournament theory was introduced by Lazear and Rosen (1981) who took sports tournaments as a starting point to study the incentive effects of tournaments. In economic terms tournaments are characterized by three particular features which have to be addressed and can be best explained with an analogy to a tennis tournament (for more details see also Lazear 1998). The application of tournament theory to academic career tournaments suggests that the larger the pay raise associated with a promotion, the higher is the level of effort candidates will exert to win this promotion. Also, academics can be assumed to mainly aim at being better in exactly those criteria that are relevant for their promotion which in most disciplines are more and more research output and publications (as is the case in the disciplines we study in our paper, namely management and economics). And it can be shown that for academic tournaments to be effective, wage gains for promotions on higher hierarchical levels have to be larger than for promotions on lower hierarchical levels; if not, incentives are distorted and output shrinks. Thus, we have three predictions on the effect of tournaments on individual output in academia. Firstly, we expect publication output to be higher in time periods before a promotion takes place and we expect output to slow down after a promotion has been received. Secondly, we expect academics to produce more research output if the ‘prize spread’ is larger, i.e. if the difference in wages between two academic career levels is larger. Thirdly, on higher hierarchical levels we only expect output increases before a promotion if the wage structure follows a convex shape, i.e. if the wage gains are larger on higher than on lower hierarchical levels; if not, output is not expected to be increased before a higher level promotion. Thus, we have the following three hypotheses which we will have to operationalize and test in the empirical part of the paper: Hypothesis 1: Output is higher in time periods before a promotion takes place and lower after a promotion has been received. Hypothesis 2: Output is higher where the difference in wage levels between two career steps is larger. Hypothesis 3: Output before a promotion on higher hierarchical levels only increases if wage gains are larger on high than on low hierarchical levels; if wage gains are similar or lower on high than on low hierarchical levels, output is not expected to be increased before a promotion.

To test our hypotheses we use differences in the promotion structure of German and US universities.

SMA_2008_Proceedings-Page0727

Institutional Settings: Career Incentives in German and US Universities In order to test our model we need information on typical career steps in both university systems and the wage gains attached to them. To get there we first give an overview on the similarities and differences in the two university systems and in a second step study in more detail the career incentives in each system. Characteristics of the German and US University System Despite some major differences in the regulatory framework of German and US universities, there are many functionally equivalent institutions and regulations as shown below. This is very important for our study because only the similarities in the overall structure of the career systems make them suitable for a comparison of the impact of career differences. At the same time there has to be enough variation in the academic career system so that we are able to study the impact of differences in career patterns on publication output. In their classic study of the academic marketplace and particularly academic careers, Caplow and McGee (1958) concluded already that the core characteristic of any academic career system is the so called “vacancy-and-replacement mechanism”. They consider this process by which a department replaces its members as central to the understanding of any academic institution. In their study they extensively analyze the inner workings of the governance mechanisms within a department, and Baruch and Hall (2004) conclude that the mechanisms revealed by Caplow and Mc Gee in the late 1950s, still sound familiar and valid in today’s environment. Caplow and Mc Gee also describe the “nitty-gritty” details of the evaluation process, the kinds of credentials that were used and the important role of peer review and reputational measures for a career decision. Since reputation or prestige is determined in a very early stage of the career (mainly by number of citations and high-quality publications) Caplow and McGee conclude that an academic’s career is more or less set by the age of 40. Baruch and Hall (2004) further argue that university systems are typically characterized by flat hierarchies, usually consisting of three steps that typically make advancement a “non-issue after the person (…) attained senior rank” (p. 248). This again is a very important aspect for our study since it has major consequences for the incentive intensity over the life-course of an academic. It means that after having attained the last step (which is usually quite early) there are no more career incentives for the rest of the time until retirement. Despite the above mentioned similarities there are also substantial differences between the two systems. The higher education system in the US is much more stratified than the German system. The US system contains of several thousands of public and private institutions. The US system is very hierarchical with its well known ivy league universities at the top and a broad array of different types of universities on lower levels. At the upper end there is a strong focus on elite and excellence, rather than on assuring common minimum standards. The German university system in contrast, is almost solely government funded and is based on the principle of equality of opportunities rather than on differentiation. In consequence, it is largely centralized and homogeneous and the number of universities is much smaller in Germany. Career Incentives in the German System In Germany, unlike in the US, a doctoral dissertation used to be only the second last step towards a researcher’s career. Students who intended to follow a university career usually stayed at their doctoral institution and were typically full time employed as senior assistants (Hochschulassistent). This is a temporary position with a maximum length of six years and no option to stay longer or even receive tenure. The assistant is assigned to a professor and teaches and works on research projects in consent

SMA_2008_Proceedings-Page0728

with the professor. During these six years the young researcher was expected to write a second thesis, the so called “Habilitation”, which was a requirement to apply for a regular professorship. However, the young researcher could not apply for a professorship at the degree-granting institution, he or she had to change places for the next career step. After finishing the Habilitation, the young researcher was on the job market competing in recruitment tournaments. All candidates that were on the market for a particular position at a given time were ranked by an internal recruiting committee according to their quality, and the job was offered to the first ranked candidate. If he or she accepted the job offer, he/she was granted tenure and received an income according to an income scale that was fixed by law on a national level. Furthermore, German professors are lifetime civil servants, they cannot be laid off and their pension is completely taken care of by the state (according to a generously defined benefit plan). On the other hand, candidates who keep failing to win a recruitment tournament have to look for employment opportunities outside the university system. Adequate employment opportunities outside the university systems are typically rare and much less attractive because job applicants with a Habilitation are often considered to be overqualified for “regular” jobs on the external labor market. Therefore, “losers” of the first recruitment tournament often cannot expect a significant wage increase but will more likely even have to accept a decline in wages because of their over-qualification. Career Incentives in the US System In order to compare the career incentives for US and German university researchers we have to look for functional equivalents in the careers of researchers in the US system. As we have seen the first step towards a career as a researcher in the German system was the position of an assistant, which is functionally equivalent to the position of an assistant professor in the US system: both just finished their dissertation, the German assistants hold a doctoral degree and the US assistant-professors a PhD. In addition, both are employed on a temporary basis for about six years and the subsequent career step is a lifetime employment (a lifetime civil servant contract in Germany and a tenured position in the US). However, there is one major difference. Assistant professors in the US are autonomous in teaching and research, whereas German assistants are still advised by tenured professors. Like in Germany, wages of US professors depend heavily on ranks. Other than in the German system, where there is one dominant career step from assistant to full professor, in the US system there are two major steps. The first is from assistant professor to associate professor with a large increase in employment security because it is tenured but a rather modest increase in income. The second step is from associate professor to full professor with a larger increase in income, particularly at research universities. In addition to the monetary gains, young researchers also gain reputation if promoted from assistant to associate professor and they gain even more reputation if promoted to full professor or even to a distinguished chair. So the short term gains of winning the first round in the academic career tournament, i.e. the promotion from assistant to associate professor, are substantial in terms of employment security, but not so much in terms of monetary gains or reputation. At the same time, the options for further gains in subsequent tournaments in terms of income increase or reputation are relatively high when compared with the German system. If we plug these institutional differences into our general hypotheses derived from tournament theory we are able to derive the following testable hypotheses. Empirically testable Hypotheses Firstly, in both countries, there is an extraordinarily large winner's prize for winning the first round of the academic career tournament due to becoming lifetime employed or tenured. The incentives are basically identical in the German and the US system. According to hypothesis 1 we expect output to be

SMA_2008_Proceedings-Page0729

higher before and lower after tenure in the US and lifetime-employment in Germany. (H1). Secondly, the prize structure in the USA and Germany differs systematically in subsequent rounds of the tournament. A promotion to a full professorship in the US yields higher income and reputation gains than a second or third call in Germany. According to hypothesis 2 we expect increased research output in the period before a promotion to full professor in the US to be higher than in the period before receiving a second or third call in Germany (H2). Thirdly, since wage gains are convex in the US system up until full-professor level, i.e. wage gains are larger if a person is promoted to the full professor than to associate professor or assistant professor, according to hypothesis 3 output should be increased before and decreased after promotion to full professors. In the US we expect research output to be higher before promotion to full professor and lower after promotion to full professor (H3a). Since wage gains for second or third calls in Germany are not convex, i.e. wage gains are not increasing but even decreasing along the academic career ladder, according to hypothesis 3 output should not be increased before such calls. In Germany we expect neither research output to increase before second or third calls nor do we expect them to decrease after such calls (H3b). Empirical findings: career incentives and publication output To test our hypotheses, we collected data from 189 German researchers and 112 US researchers in management and economics. Research output was measured by a weighted number of publications in journals and edited books. As a dependent variable in our regression analysis we use individually standardized publication indicator (OUTPUTt,i or OUTPUTWEIGHt,i). We use standardized outputs because in order to test our hypotheses we are only interested in the shape of the lifecycle publication profile but not in the absolute level of a researcher’s lifecycle publication output because our hypotheses only predict individual changes in output over an individual’s academic career cycle. According to our theoretical explanation the most important explanatory variables are different career steps in Germany and the US. To test our implications we estimate separate linear regression models for US and German researchers with OUTPUTt,i in period (t) as a dependent variable and various career steps (promotions) in periods (t-1), (t) or (t+1) as explanatory variables. Our results show that all three hypotheses regarding the impact of career incentives on publication output are borne out well in the data. As expected in H1 research output is significantly higher in the period before a promotion takes place (positive sign of PROMOTIONt+1,i) and it is often below average after a promotion took place (negative sign of PROMOTIONt-1,i; cf. model I, US and Germany). So these results clearly indicate that researchers strongly and instantly adjust their publication effort as a reaction to the existence or non-existence of career events and the monetary rewards attached to them. If we split PROMOTION into “tenure decision” and “promotion to full professor” for US researchers (model II, US) and into first appointments and re-appointments for German researchers (model II, Germany), we find that in the US both types of promotions have almost the same impact on weighted publication output OUTPUTWEIGH but not for German researchers as expected in H2. In the US we observe a significantly higher research output in the period before tenure and before full professor takes place. In Germany however we only observe a significantly higher research output in the period before the first appointment but not in the period before re-appointments. According to tournament theory this is because in Germany potential gains in the first round of the promotion tournament are substantially higher than in later rounds although tournament theory would suggests that gains in later rounds should always be higher than in earlier rounds to keep tournament incentives intact.

SMA_2008_Proceedings-Page0730

As expected in H3 we find that in the US output is below average after the promotion to full professor (FULLt-1 is significantly negative in US model II OUTPUTWEIGH). Conclusions Taking the results of Melguizo and Strober (2007) and our results together we conclude that increasing dominance of prestige oriented reward systems enforce and accelerate the well-known publish or perish process. Since all types of higher education institutions are more and more emulating research institutions in their pursuit of prestige and publication output and monetary rewards for such prestige maximizing research activities will become ever stronger faculty members are induced to neglect important parts of academic activities such as teaching, advising and community services. As a consequence and if nothing changes this process causes university institutions to drift away more and more from parts of the higher education mission.

References Baruch, Yehuda, Hall, Douglas T. (2004): The Academic Career: A Model for Future Careers in Other Sectors? In: Journal of Vocational Behavior 64(2004), 241-262. Caplow, T., McGee, R.J. (1958): The Academic Marketplace. Garden City, NY: Anchor Books. Lazear, Edward P., Rosen, Sherwin (1981): Rank-Order Tournaments as Optimal Labor Contracts. In: Journal of Political Economy 89(1981)5, 841-864. Melguizo, Tatiana, Strober, Myra H. (2007): Faculty Salaries and the Maximization of Prestige. In: Research in Higher Education, 48(2007)6, 633-668.