âMatthew Effectsâ or Governance Effects?

“Matthew Effects” or Governance Effects? An Analysis of Performance‐Based Funding in German University Medicine Paper presented in track 8 at the EAIR 33rd Annual Forum in Warsaw, Poland 28‐31 August 2011 Name of Author(s) René Krempkow Uta Landrock Contact Details Dr. René Krempkow Institute for Research Information and Quality Assurance Godesberger Allee 90 53175 Bonn GERMANY E‐mail: [email protected] Key words policy research, institutional performance measures, funding of state higher education institutions, organisational structures, governance

1

Krempkow & Landrock: “Matthew Effects” or Governance Effects? An Analysis of Performance Based Funding …

Abstract “Matthew Effects” or Governance Effects? An Analysis of Performance‐Based Funding in German University Medicine In recent years, performance‐based funding (PBF) has been central among competitive elements in universities. Faculties across Germany have also been adapting to PBF, but empirical findings thus far provide no well‐defined answers about its impact. This ambivalence often leads to citing “Matthew effects” as unintended consequences of how third‐party funding is governed. Conversely, the volume of third‐party funding from institutions is seen as evidence of successful governance. We employ results of a multivariate analysis in hopes to find the extent to which the scale of faculties and characteristics of governance relate to the volume of third‐party funding

2


Presentation “Matthew Effects” or Governance Effects? An Analysis of Performance‐Based Funding in German University Medicine 1 In recent years, performance‐based funding (PBF) has taken a central role among competitive elements in German universities. Links between performance evaluation and mechanisms of financial distribution are now in place not only at statewide levels, but also at faculty levels. In this article, we empirically examine the extent to which internal faculty PBF has intended and non‐intended effects on the volume of third‐ party funding and publication output. Results of a first analysis indeed show that researched characteristics do not accompany intended effects regarding the case of medicine in Germany. However, with a broader consideration of overall governance2, intended effects are definitely found. 1. Assumptions about Effects of Control and Governance The main ideas of the New Public Management (NPM) framework, as well as New Governance literature, to some extent, are based on the assumption that output orientated governance is the most efficient form of governance. Actors in ministries and universities, such as employees and deans who are concerned with the conception of PBF models, also assume that a stronger weight of third‐party funding criteria in the PBF formula should lead to raising a higher level of third‐party funds. They attempt to influence a different arrangement of PBF from supposed or actual performance deficits. In contrast, other actors, such as some from the sociology of science, foster fundamental doubts as to whether a control of science is possible in this way. As of yet, however, little is known about the impact of PBF as a means of governance – neither from a national nor from an international perspective (see Butler 2010). Also, recent empirical findings do not offer clear answers about whether such a method of governance actually leads to increased performance, and to what extent unintended effects occur. On one hand, so‐called “Matthew effects” (“to all who have, more will be given”) are seen as unintended outcomes of governance based on third‐ party funding (for example, see Jansen et al., 2007; Zechlin, 2008; Münch, 2008; Jansen et al. 20093). On the other hand, the increased volume of third‐party funding from institutions is cited as proof of successful governance (for example, see Jäger, 2008; Auspurg et al., 2008; Hilzenbecher, 2010). Despite this research, comparative effects of governance and the impact of Matthew effects on third‐party funding are rarely investigated empirically (Butler, 2010). An analysis of German university medicine suits this question well, given that it has been over a decade since the first PBF models were implemented nationwide in medical faculties. The Institute for Research Information and Quality Assurance (iFQ) in Bonn therefore investigates this problem through a project about the organisation, perception, and effects of performance‐based funding.4 It builds on the framework of our iFQ project that conduced a wide spectrum of methods: guided interviews, document analyses of PBF models (e.g. model descriptions), analyses of statistical data from university medicine, a survey of faculties5, internet research and – in future steps – bibliometric analyses. 2. third‐Party Funding Analyses The aim of the following analyses is to investigate both intended and unintended effects of PBF systems. We will present selected results of a multivariate analysis of PBF in relation to the volume of third‐party funding of medical faculties (spent third‐party funds per staffed professor 2003‐2005). The hypotheses builds on the framework of our iFQ project that conduced guided interviews with deans and/or deans of research (Forschungsdekane), and furthermore document analyses of PBF models. We will begin by considering characteristics of PBF models (i.e. performance criteria and their weighting) and the implementation of PBF (when they were initiated, evaluated, and revised) along side the structural characteristics of the faculties’ intensity of publications, and initial financial conditions. Additionally, we will review the relationships between other structural characteristics of the faculties, such as the deans’ scope of decision‐making or the presence of vice deans of research. The following Figure 1 describes our model.

3


Figure 1: Analysis model of medical faculties’ volume of third‐party funding

Faculty Governance

Publication Activity

e.g. PBFimplementation

e.g. Budget

e.g.: publications per prof.

Research Input

Core processes

Output

(Teaching)

structure e.g.: Deans’ term period

Graphic: our own depiction (following Nickel, 2007; Teichler, 2003) 4


here: 3rd party funding per prof.

We have depicted the central hypotheses in the following outline (Figure 2). Figure 2: Hypotheses for correlations between PBF and volume of third‐party funding Model characteristics of internal faculty PBF: A higher weight of PBF criteria (e.g. third‐party funding) correlates positively with later performance in this area (here: third‐party funding). PBF systems that are based on evaluations, that have been introduced for a longer period of time and therefore probably are more strongly established, and that have been altered (and therefore probably further developed) accompany higher performance. Interaction with other dimensions of performance: A high performance of publication correlates positively with performance of third‐party funding (and vice versa). Structural characteristics of the faculties: Longer term periods of deans (as proxy indicators of the establishment and esteem of this function) are accompanied by a higher performance of the concerned faculties. Input / initial conditions: Higher investments and federal allocation of sums (LZB) to the concerned faculties are useful for research performance (here: third‐party funding). Contrary to the expectations, and also to the statements from PBF actors in ministries (see, for example, Hilzenbecher 2010), the results of our multivariate analysis established that the actual volume of third‐ party funding per professor and the researched characteristics of PBF models are not related. The researched characteristics of PBF models were the weighting of third‐party funding (as a performance criteria) and the implementation characteristics of PBF (when they were initiated, evaluated, and revised). However, a comprehensive evaluation process, the size of financial resources, and intensity of publications are related to third‐party funding. Also significant is the connection between the volume of third‐party funding and deans with long tenure, although this is not as strong as the aforementioned relationships. The results of our model calculations prove to be stable. We will favour Model 4 because of its corrected R2 of 0.61, acceptable multicollinearity, and highest explanatory power while applying variables sparingly.6 The following Table 1 shows the strength of the found connections.7 Table 1: Standardised beta‐coefficients for regression models with the dependent variable Spent third‐ party funds per staffed professor (in thousands €) 2003‐2005 Model 2 Model 3 Model 1 (corr. R2 Variables: (corr. R2 (corr. R2 =.57***) =.59***) =.61***) Weight of third‐party funds in PBF research (%) ‐.10 ‐.09 ‐ Implementation period (before 2000=1, from=0)8 .12 .13 .13 Revisions in PBF from 2004 (yes=1, no=0) .05 ‐ ‐ ‐.52*** ‐.50*** Funding allocation of depart. based on evaluation ‐.50*** procedure? (1=research, 0=research+teaching)9 Publications per scientific staff 2003‐05 (num.) .43** .52*** .44*** Term period of deans (in years) .29 .30** .26** Total budget ‘03‐05, central bank + invest (in €) .41** .44** .41***

Model 4 (corr. R2 =.61***) ‐ ‐ ‐ ‐.48*** .43*** .28** .38***

Data sources: Landkarte Hochschulmedizin, 2007; Brähler, 2009; self‐made inquires, 2010

5


In further model variations (which are not depicted here), we have examined the extent to which the results change through the inclusion of further structural characteristics (cooperation vs. integration model10, availability of deans of research). Furthermore, a separate inclusion of investments and federally allocated sum occurred. All models verified the substantial connections regarding evaluation proceedings, publication activity, and the provision of resources. Only the term period of deans show partially smaller and insignificant effects.11 Therefore, further analyses also point to results that find the strongest effects of PBF based on evaluations and publication activities. The result that larger publication activities are accompanied by higher third‐party funds is expectable and corresponds with results of other studies. Effects of PBF’s basis of evaluation were also expected. However, they initially find themselves here in an unexpected direction, because they do not accompany the allocation of third‐party funds based on evaluation of research. In fact, it is an allocation of funds based on more extensive evaluations of research and teaching. Conversations with heads of faculty and administration in medical faculties suggest two possible interpretations of this. First, it could be that a larger number of individual evaluations (separately for research, teaching, promotion of young associates, etc.) are unfavourable against an overall concept of evaluations that is more extensive and balanced, such as it has been increasingly demanded in recent times. This would also correspond to the partial criticism toward “evaluitis”12 at universities (Frey, 2008), which, among other things, aim at (too) large a number of unbalanced individual measures. Second, it could also be that the variable of evaluation procedures recorded here is not the “true” reason, and is instead only an expression of an underlying, more abstract dimension. Possibly, this could be the faculties’ strategic ability: because an overall concept of evaluations that is more extensive and balanced would be conceivable as a part of an overall strategy to develop the faculty.13 These must remain initial assumptions for now, and are to be examined in further research, which would include qualitative analyses. The positive effect of deans’ longer term periods on the volume of third‐party funding corresponds with a clear increase with the terms’ average lengths. By now, this accounts for deans at medical faculties who considerably more often hold full‐time offices than before. This infers a clear higher attractiveness to the office, as reported from other disciplines at German universities. Together with the deans’ decision competencies that were expanded in the course of introducing NPM, this could account for the strategic abilities of the faculty and / or its management.14 It appears more necessary to interpret that the weight of the third‐party funds in the PBF model, just as with the introduction time and modifications in the PBF, are not connected with current volume of third‐party funding. However, it does not have to inevitably mean that the weight is irrelevant. Our results could also be at least partially attributed to that fact that some PBF models provide for capped limits (Kappungsgrenzen) (e.g. in Baden‐Württemberg, see Krempkow, 2010), or that special provisions for a part of the professors are in force (e.g. protection of existing provisions on the basis of still applicable appointments, or a similar situation). We nonetheless presume from our existing in‐depth analyses of selected PBF models that such limits and special provisions, even with a pessimistic view, should not lead to the fact that any (potential) effect of redistribution will be annulled.15 International experiences show moreover that PBF, with its indicators and relative weight, even with relatively small redistributed sums, can have the potential to turn into a strong driving force for institutional priorities, indeed even beyond immanent comparative tables and their discussion (see Harris, 2007). The precondition for this is that – as it is usual in university medicine – PBF criteria and results are well‐known. Here it seems overall that further research is necessary in order to interpret and understand not only the results according to the weight of third‐party funding, but also according to the point in time in which PBF was introduced, and any alterations in PBF. Overall, our results shows that the relations between the investigated characteristics of PBF models are too complex for direct effects of – for example – the weighting of third‐party funding in the PBF models. In any case, PBF does not accompany the expected effects, at least as they relate to German university medicine. In contrast, our observations of faculty governance (also independent from concrete properties of the PBF models) have indeed found intended effects. Then again, in addition to such effects of governance, as well as the expected impact of publication intensity and effects of initial financial conditions, there is also evidence of Matthew effects, which is an outcome that some sceptics of PBF had expected to see. However, these Matthew effects are not contrary to the predictions, but only stand “in moderation” (Hornbostel/Heise, 2006, p. 25) to the volume of third‐party funding. 6


3. Publication Analyses Other than our analyses of third‐party funding, we have also carried out publication analyses. Our goal is to figure out which forms of PBF models and characteristics of medical faculties have effects on the volume of publication. Publications are highly relevant because, along with third‐party funding, they serve as an important scale of scientific performance. In this respect, they are targets of NPM’s control of output. The basis of the publication analyses is the already depicted third‐party funding model, in which some adjustments of content are necessary: It must be verified whether the extent to which the PBF and structural characteristics which were applied to third‐party funding analyses as well as initial conditions can also be relevant for publications. To test for this, the basic model uses the number of publications as an explanatory variable instead of third‐party funding as a dependent variable. In the analogous application of this model, the spent third‐party funds per professor from 2003‐2005 will be incorporated as an independent variable and a further dimension of performance. We did not decide on this approach only because there are correlations between third‐party funding and publications16, but also because of the specific double aspect of third‐party funds: As a result of successful research, third‐party funds are not only output variables, but they also can be regarded as input variables (see Hornbostel/Heise, 2006; Jansen et al, 2007). It can be assumed that procured and spent third‐party funds increase research activities and thus also the output of publications. An important modification is that the weight of the independent variable from the third‐party funding in the PBF is replaced through the variable weight from publications in the PBF. We assume a presumption within medical faculties that a higher weight of publications leads to a stronger performance in this area. As an extra independent variable, the impact factor at the time of appointment is taken into consideration. Our hypothesis here is: A higher publication output occurs when impact factors (IF)17 are already considered during the appointment of full‐time professors. It is presumed in the faculties that those with higher impact factors tend to be responsible for a higher number of publications. The dependent variable in the analyses presented here is the number of publications with peer‐review from 2006‐2008 per professor (as a three year average). We decided for this because, at this point, we assume that changes in the volume of publications are not only a chronological trailing effect of control stimuli, but also of the volume of third‐party funding.18 Our theoretical model also contains the possibility of feedback, which should be exhausted in the pending analyses about the interdependences of input and output factors. The basic model to explain the number of peer‐reviewed publications from 2006‐2008 per professor shows highly significant results and accounts for almost half the variance of the examined performance dimensions in Figure 4 with a corrected R2 from .49 in Table 2: Table 2: Standardised beta‐coefficients for regression models with the dependent variable publications with peer review per professor, 2006‐2008 Model 2 Model 3 Model 4 Model 1 (corr. R2 (corr. R2 Variables: (corr. R2 (corr. R2 =.46**) =.48***) =.49***) =.49***) Weight of third‐party funds in PBF research (%) ‐.48** ‐.50** ‐.51** ‐.46** Implementation period (before 2000=1, from=0) ‐.19 ‐.18 ‐.20 ‐.20 Revisions in PBF from 2004 (yes=1, no=0) ‐.35* ‐.34* ‐.40** ‐.42** Impact factor during appointment? (yes=1, no=0) .26 .27 .30* .31* Spent third‐party funds ‘03‐05 per staffed prof. (in .55** .51** .48** .40** thousands €) Funding allocation of depart. based on evaluation .17 .18 .18 ‐ procedure? (1=research, 0=research+teaching Term period of deans (in years) ‐.15 ‐.15 ‐ ‐ Total budget ‘03‐05, central bank + invest (in €) ‐.07 ‐ ‐ ‐ Sources: Landkarte Hochschulmedizin, 2007, 2010; Brähler, 2009 self‐made inquires, 2010

Arranged after the strength of influence, the statistically significant variables prove to be: the weight of the publications in the research PBF, alterations in the PBF, the third‐party funding per full professor from 7


2003 until 2005 and the consideration of the impact factor during appointments of professors. This amounts to the following correlations: Contrary to the assumptions, a lower weight of publications in the research PBF accompanies a higher number of publications per professor. As formulated in our hypotheses, the amount of spent third‐party funds per professor correlates positively with the volume of publications per professor, but considerable alternations in faculty PBF is accompanied by a decreasing volume of publications. The consideration of impact factors during appointments correlates positively with the number of publications per professor. The time period in which PBF was introduced has no independent significant effect on the number of publications. Therefore, there is not a provable and empirical connection between these PBF characteristics and the output of publications. Nevertheless, the time of introduction is at least indirectly effective, because the removal of this variable from the model would slightly lower the explanatory power of the entire model. In contrast to the volume of third‐party funding, the entire budget provides no independent explanation for this. In summing up the publication analyses, is can thus be ascertained that our hypotheses regarding the volume of third‐party funding and the consideration of impact factors during appointments have been confirmed. Although impact factors are not a measurement for an individual scientist’s performance competence, our models still allow for the assumption that publication performance will be promoted in faculties that consider the impact factor during appointments. The result that alterations of the PBF models accompany a reduced number of publications might indicate that stable PBF systems have more positive effects on publication performance than systems that are likely to be subjected to changes and modifications. An alternative explanation would be that in this dimension of performance, unsuccessful faculties undertake more efforts – including a higher weight of publications as well as alterations in the PBF systems – but they nonetheless achieve fewer publications for other reasons. The result that higher weights in the PBF accompany a fewer number of publications per professor does not correspond with our hypotheses and might indicate unintended effects. In contrast to the already depicted model of third‐party funding, our models show that different characteristics of governance display different effects for both performance indicators of third‐party funding and publications. For obtaining third‐party funding: publication activity, evaluation procedures, and the entire budget showed the most effects. Accordingly, the most effects for the output of publications were brought about by the weight in the research PBF. An effect of the entire budget on the level of third‐party funding is not ascertainable for the volume of publications. However, as expected, interdependences between the level of third‐party funding and the number of publications are evident: an increase of one accompanies an increase of the other. 4. Outlook In the separate consideration of both performance indicators, our analyses of the third‐party funding and publication performances already suggest that the combination of PBF characteristics under consideration of structural and initial conditions is very complex, and at least no direct controlling effects coming from a higher weight of certain indicators are provable. Medical faculties face challenges when attempting to control both output dimensions because unintended effects can also increasingly occur along with intended effects. Our goal in further analyses is to better understand the complexity of this combination. The coming months should yield successful bibliometric analyses that are analogous to the study of medical faculties’ volume of third‐party funding and publication that are depicted here. Moreover, an assessment of PBF via individual scientists in university medicine is still missing from an extensive, multi‐ perspective consideration of PBF effects. This would be additional to the already compiled perspectives found in expert interviews and document analyses. For this, researchers conducted a standardised online survey in summer 2011, which is currently being analysed. In order to find out under which conditions PBF is relevant to action, we asked the researches how they assess the PBF models existing at the time of their study, and to what extent they orientated their actions toward them. Here, we also want to identify actors’ working conditions, motives, interests, and publication strategies, in order to understand dynamics of change, when necessary. Moreover, we want to examine the roles played by perception in relation to the meritocratic justice (Leistungsgerechtigkeit) of PBF systems. In doing so, we also hope to learn more about the reasons behind the partially unexpected results of our regression analyses. 8


The experiences with PBF, both positive and negative, could also be interesting for (medical) faculties outside of Germany. With our results, we hope to answer questions about the relation between the scale of faculties and characteristics of governance and the volume of third‐party funding. We hope to find out which possibilities ensure that PBF models can support scientists’ research activities and manifest the least possible unintended consequences in the future. Notes 1 This paper is based on a presentation from the 6th Annual Conference of the German‐speaking Higher Education Research Society in May 2011 in Wittenberg, and the conference of the working group “Sociology of Organisations” in the German Sociology Association in June 2011 in Dortmund. 2 Here, we support a relatively ready wide understanding of governance as perspective, in which control is included as a part of governance (compare with Mayntz, 2005; also more detailed in Schulz, 2010). 3 Beyond that, Görtz et al. (2010) try to find the effects of successful implementation(s) of PBF on open outcomes research in astrophysics, nanoscience, and the economy. 4 Our investigation is promoted by the Federal Ministry of Education and Research. 5 For kindly providing the data, we would like to give sincere thanks to Elmar Brähler, Universität Leipzig. For a description of the survey, see Brähler/Strauss (2009). 6 We previously tested the approximate normal distribution of the dependent variables (by Kolmogorov‐Smirnov test/ Shapiro‐ Wilk test and Q‐Q plot), which is given. 7 The corrected coefficient of determination in the first row of the table stands as a measurement for the explanatory power of the model. The standard beta‐coefficients in the following rows are a comparable measurement of the individual variables’ explanatory power. In all cases: the nearer to value 1, the more convincing the results. The marks ***/**/* behind the numerical values mean significance on the 1‐ / 5‐/ 10‐ percent alpha error level. (Statistical significance is not strictly necessary for complete sampling, but it is common.) 8 The median split was applied for dichotomisation in order to achieve similar group sizes. 9 It was initially expected that a basis of evaluation has positive effects, as opposed to no such basis. All faculties indicated that their funding allocations are based on evaluation procedures. One part was based on evaluation procedures for research only; the other part was based on evaluation procedures for research and teaching. 10 The cooperation model means that two separate institutions have to cooperate (a faculty of medicine and a university hospital at the same location). The integration model means that a faculty of medicine and a university hospital at the same location are integrated in one institution. 11 Moreover, further models that include the spent third‐party funds per research associate (in thousands of €) achieves an 2 even a higher power of explanation (corr. R =.79). As before, the same three variables show the highest beta‐coefficients (entire budget 2003‐2005 based on the allocation of funds on an evaluations procedure, publication per research associate.) We decided to analyse the spent third‐party funds per professor because they are mostly used as indicator in the PBF. 12 That means too many evaluations. 13 This also agrees with the below mentioned connection with deans’ term periods and volume of third‐party funding. With a longer term period, it is considered easier to develop and realise more comprehensive strategies (Scholkmann et al., 2008). 14 König (2011) points out – referring to Authur Benz – that the capability of strategically controlling the reform goals belongs to almost all federal states. The prerequisite is that those in charge of the universities’ managerial levels are actually in the position to develop strategic perspectives. 15 Moreover, there have been relatively large sums of PBF distributed in many federal states for a few years now (Krempkow, 2010 for full details). Other authors who examine effects of selected PBF models of federal states assume larger distributed sums or shares of distribution from larger effects of control (see König, 2011, and other authors quoted therein). But König also points out here that the impact of PBF on the concrete practice at universities until now has been rarely documented. 16 The Pearson correlation coefficient between the third‐party funding per full professor from 2003‐2005 and the publications per professor from 2006‐2008 amounts to 0.35**. 17 The impact factor of a journal measures how often its articles are cited in other scientific journals. However, “for the assessment of the scientists’ performance in terms of publications, the journal impact factors are indeed not suitable” (Lewandowski 2006). Nevertheless, the factors were applied as such to 27 (from 36) medical faculties. rd 18 For that reason, we have not used here the results from already carried out analyses about 3 party funding from 2006‐2008. The analyses of third‐party funding for the years of 2006‐2008 show relatively similar results to those for 2003‐2005.

Acknowledgements We gratefully acknowledge the cooperation of our colleagues Jörg Neufeld, Patricia Schulz und Verena Walter. We are also indebted to Jeff Purchla for his advice and support for translation. We like to thank the discussants of the 6th Annual Conference of the German‐speaking Higher Education Research Society, and the conference of the working group “Sociology of Organisations” in the German Sociology Association for their suggestions. Of course, we also thank the organisers for creating the possibility for us to discuss our research. 9


References Auspurg, K., Hinz, T., Güdler, J. (2008): Herausbildung einer akademischen Elite? Zum Einfluss der Größe und Reputation von Universitäten auf Forschungsförderung. In: Kölner Zeitschrift für Soziologie und Sozialpsychologie (KZfSS) 4/2008, 653‐685; Brähler, E. & Strauß, B. (2009): Leistungsorientierte Mittelvergabe an Medizinischen Fakultäten. Eine aktuelle Übersicht. In: Bundesgesundheitsblatt 9/2009, 910‐916; Butler, L. (2010): Impacts of Performance‐Based Research Funding systems: A review of the concerns and the evidence. In: Performance‐Based Funding for public research in tertiary education institutions. workshop proceedings. OECD publishing 2010: 127‐166 Frey, B. S. (2008): Evaluitis ‐ eine neue Krankheit. In: Matthies, Hildegard/Simon, Dagmar (Hrsg.): Wissenschaft unter Beobachtung: Effekte und Defekte von Evaluationen. Wiesbaden: VS Verlag für Sozialwissenschaften, 124‐140. Görtz, R. von, Heidler, R., Jansen, D. (2010): Chancen für neue Forschungslinien? Leistungsorientierte Mittelvergabe und „ergebnisoffene“ Forschung. In: Beiträge zur Hochschulforschung 2/2010, 8‐32. Harris, K.‐L. (2007): A critical examination of a recent performance‐based incentive fund for teaching excellence in Australia. In: Longden, B./ Harris, K.‐L.: Funding Higher Education: A Question of Who pays? EAIR‐Monograph 2, Amsterdam, 62‐78. Hilzenbecher, M. (2010): Leistungsorientierte Mittelvergabe (LOM) in der Medizin in Baden‐ Württemberg. Genese, Bilanz und Perspektiven. In: Das Krankenhaus 7/2010, 652‐662; Hornbostel, S./Heise, S. (2006): Die Rolle von Drittmitteln in der Steuerung von Hochschulen, in: Berthold, C. et al. (Hg.), Handbuch Wissenschaftsfinanzierung, Berlin: Raabe, (part B 1.1) 1‐26. Jansen, D., Wald, A., Franke, K., Schmoch, U., Schubert, T. (2007): Drittmittel als Performanzindikator der Wissenschaftlichen Forschung. Zum Einfluss der Rahmenbedingungen auf Forschungsleistung. In: KZfSS 1/2007, 125‐149; Jansen, D., Heidler, R., Görtz, R. von (2009): Ungleiche Chancen im Wissenschaftssystem: Artefakt oder Realität? In: KZfSS (61), 363 – 467; König, K. (2011): Hochschulsteuerung, in: Peer Pasternack (Hg.), Hochschulen nach der Föderalismusreform, Leipzig: Akademische Verlagsanstalt, Leipzig (in print). Krempkow, R. (2010): Performance Based Funding: First effects of local incentive programs on the example of the German university medicine. 7th International Workshop on Higher Education Reform (HER), Reform of University Governance ‐ Trends, Policies, Fads, and Experience in Comparative Perspective, Centre for Policy Studies in Higher Education and Training (CHET) and The University of British Columbia (UBC), 07.‐08.10.2010, Vancouver. Lewandowski, D. (2006): Journal Impact Faktor, iQ‐Text: www.forschungsinfo.de/iq/agora/Journal_Impact_Factor/journal_impact_factor.asp Mayntz, R. (2005): Governance Theory als fortentwickelte Steuerungstheorie? In: Schuppert, G. F. (Hg.): Governance‐Forschung. Vergewisserung über Stand und Entwicklungslinien. Baden‐Baden: Nomos, 11‐20. Münch, R. (2008): Die Schattenseite der Errichtung von institutionellen Leuchttürmen in der Wissenschaft: Wie Konzentrationsprozesse die Produktivität pro Personaleinsatz verringern. In: iFQ Working Paper No 4, 59‐68; Nickel, S. (2007): Institutionelle QM‐Systeme in Universitäten und Fachhochschulen. Konzepte ‐ Instrumente ‐ Umsetzung. Eine empirische Studie, CHE‐Arbeitspapier Nr. 94, Gütersloh. Scholkmann, A., Roters, B., Ricken, J., Höcker, M. (Hg.) (2008): Hochschulforschung und Hochschulmanagement im Dialog. Zur Praxisrelevanz empirischer Forschung über die Hochschule. Münster: Waxmann. Schulz, P. (2010): Die analytische Governanceperspektive ‐ Diskussion und Einsatzmöglichkeiten. IQ Beitrag: www.forschungsinfo.de/iq/agora/Governance/Governance.asp Teichler, U. (2003): Die Entstehung eines superkomplexen Systems der Qualitätsbewertung. Ein Beitrag aus Sicht der Hochschulforschung. In: hochschule innovativ Juni 2003, 5‐6. Zechlin, L. (2008): Die Zeitstruktur leistungsorientierter Mittelverteilungssysteme und ihre strategischen Auswirkungen in den Hochschulen. In: Zeitschrift für Hochschulentwicklung 1/2008, 1‐15. 10


âMatthew Effectsâ or Governance Effects?

âMatthew Effectsâ or Governance Effects?

âMatthew Effectsâ or Governance Effects?

âMatthew Effectsâ or Governance Effects?