Incentivising interdisciplinary research collaboration: evidence from Australia

ABSTRACT

There is a popular perception that interdisciplinary research collaboration can yield benefits to knowledge production, from improvements to creativity to advances in addressing real-world problems. However, studies into interdisciplinarity frequently point to material obstacles, such as burdensome time and resource requirements, difficulties in publishing, and scarce opportunities for grant funding, which imply that incentives or rewards might be required to motivate academics to collaborate with colleagues across research disciplines. This article reports the results of a survey of academics and interviews with senior university leaders at a large research-intensive university in Australia, which indicate a major difference of opinion on how to incentivise interdisciplinary collaboration. While survey respondents largely emphasised material concerns, university leaders cited cultural change as a preferred strategy. If interdisciplinary research collaboration is desired, this gap between the perceptions of academics, who do the research, and university leaders, who control the resources, will need to be bridged.

KEYWORDS:

• interdisciplinarity
• interdisciplinary research collaboration
• multidisciplinary research
• cross-disciplinary research
• transdisciplinary research
• Australia

Introduction

While single-discipline research will likely remain an important standard of knowledge creation, many benefits have also been attributed to interdisciplinary research, such as its potential for fostering innovation (Edmondson & Harvey, 2018), facilitating creativity (Oddane et al., 2015), and supporting research impact (Gooch et al., 2017), among others. According to one popular perspective, knowledge arising from interdisciplinary research is required for addressing real-world social and policy problems, which are themselves inherently interdisciplinary (Krohn, 2010; Pedersen, 2016, p. 2; Siedlok et al., 2015, p. 98). It may not be surprising, then, that interdisciplinary collaboration is often seen by university leaders as desirable, and consequently they are actively promoting and encouraging interdisciplinarity at research institutions around the world (Frickel et al., 2016, pp. 5–6; Jacobs, 2014, pp. 1–2).

However, numerous obstacles to achieving interdisciplinary collaboration have been noted. For instance, many observers have described communication challenges: research disciplines can employ discipline-specific ‘jargon’ that requires out-group collaborators to learn a new language (Austin et al., 2008, pp. 562–563; Gibson et al., 2019, p. 58; McCoy & Gardner, 2012, p. 46). Disciplines are often said to favour particular methodologies, values, and epistemologies (Brister, 2016; Lach, 2014). Disciplines may have differences in norms and strategies for publication, especially between the arts and sciences (O’Cathain et al., 2008, p. 1582). In some cases, in-group versus out-group attitudes can border on ‘tribalism’, resulting in protectionist practices to guard against encroachment on job openings or PhD student positions (Siedlok & Hibbert, 2014, p. 204; Turner, 2017). Collectively, obstacles that relate to language, values, norms, and attitudes are usually referred to as ‘cultural’ barriers to interdisciplinary collaboration (Lattuca, 2001).

In addition to cultural barriers, three major material concerns – that is, concerns related to more tangible factors, such as time, money, and other kinds of resources – are also frequently discussed in relation to interdisciplinarity. For instance, interdisciplinary research is supposed to require more resources, including time, than is needed for single-discipline research (Austin et al., 2008, p. 559; McCoy & Gardner, 2012, p. 46; Pfirman & Martin, 2017); it may be harder to get grant funding for interdisciplinary research than it is for single-discipline research (Bammer, 2016; Kandiko & Blackmore, 2008, p. 91), possibly because of difficulties in identifying experts who can assess applications that cross disciplinary boundaries (Lamont et al., 2006); and it is said to be harder to publish the results of interdisciplinary research in high quality journals (Campbell, 2005). This last concern is material in nature because academic publication records are usually tied to career advancement, including tenure and promotion.

Although the material concerns mentioned above have been discussed widely, empirical evidence to support these claims is still fairly thin (Newman, 2023), and as a consequence, the assertions in this scholarship are often disputed. Brown et al. (2015, p. 317), for instance, argue that winning grant funding for interdisciplinary research may actually be easier than it is for single-discipline research, because of the trendiness of interdisciplinarity and the consequent proliferation of interdisciplinary funding opportunities. In another example, Abramo et al. (2017) found that for some combinations of research areas, interdisciplinary research can result in higher impact publications than comparable single-discipline research. And in a counterpoint to the notion that interdisciplinary research takes up more time and resources than single-discipline research, Millar (2013) found that for early career researchers, interdisciplinary research can yield a higher number of publications than single-discipline research would in the same amount of time.

Do academics see material concerns, such as resource needs, a lack of funding opportunities, and difficulties in publishing, as barriers to interdisciplinary research collaboration? Academics are, after all, the ones who do the research, and their beliefs on these issues will affect their willingness to collaborate. Moreover, the existence of material obstacles suggests that material solutions – perhaps in the form of incentives or rewards – might be required to encourage or promote interdisciplinarity. Therefore, the beliefs of university leaders and administrators, who control many of the resources related to research, are an important consideration as these beliefs will affect the kinds of incentives and rewards that are available. Most importantly, if there is disagreement between these two groups, it may be difficult to pursue any strategy to promote interdisciplinary collaboration at all. More generally, disagreement between research workers (i.e., academics) and research managers (i.e., university leaders) may be symptomatic of a wider conflict between workers and managers in the university sector (Kenny, 2018).

This study seeks to address these questions related to material concerns around interdisciplinary research collaboration. Using Monash University, a large research-focused higher education institution in Australia, as a case study, I surveyed academics to ask them about their perceptions of material barriers to interdisciplinary research collaboration, and what they believe might incentivise them to overcome these barriers and participate in interdisciplinary research. Around the same time period, I interviewed senior leaders at the university, including faculty deans, the Provost, and the Vice-Chancellor (equivalent to President of the university), to ask them similar questions about their perceptions surrounding interdisciplinary research and how academics can be motivated to collaborate across disciplines.

In this study, as will be elaborated below, academics and university leaders did not agree on how to incentivise interdisciplinary research collaboration. Academics on average claimed to believe in the traditional material concerns about interdisciplinary collaboration discussed above, and largely pointed to material rewards and incentives as being the best way to motivate them to collaborate. University leaders also claimed to believe in the same material concerns, but almost uniformly agreed that material rewards or incentives were not productive strategies, referring instead to ‘cultural change’ as the best way to encourage interdisciplinary collaboration.

Ultimately, if interdisciplinary collaboration is to be achieved, the people who do the research work and the people who control the research-related resources will have to agree on what the obstacles are and how to overcome them. Understanding where disagreements may occur between these two groups is a first step towards designing effective strategies for promoting interdisciplinary collaboration.

Interdisciplinary research and collaboration

Interdisciplinarity is a somewhat contentious term, and there is a wide literature attempting to define it, with numerous models and typologies on offer (Aboelela et al., 2007; Klein, 2017; Newman, 2023). The standard definition of interdisciplinarity put forward by the US National Academies is ‘research by teams or individuals that integrates information, data, techniques, tools, perspectives, concepts, and/or theories from two or more disciplines or bodies of specialized knowledge’ (Institute of Medicine, 2005, p. 26). However, not satisfied with this definition, many authors have argued that there is actually a spectrum of activity ranging from ‘multidisciplinarity’, in which disciplines or discipline-based researchers interact in parallel, without much mixing; to ‘interdisciplinarity’, where disciplinary borders are crossed or blurred but where the original disciplines are still apparent; to ‘transdisciplinarity’, where fundamentally new disciplines are created, as in biochemistry or economic history (Austin et al., 2008; Bark et al., 2016; Barry et al., 2008; Gibson et al., 2019).

Still, much definitional precision is lacking. The boundaries separating multidisciplinarity from interdisciplinarity and transdisciplinarity are highly subjective, and these terms are often used interchangeably (Bammer, 2016, p. 2). The term transdisciplinarity has in recent years taken on a new meaning – research that results from cooperation with end-users (e.g., Frodeman, 2011) – and it is not entirely clear whether transdisciplinarity under this definition implies the use of knowledge from multiple research disciplines or whether work in a single knowledge area, say, occupational therapy, could be transdisciplinary on its own by virtue of the interaction of the researchers and end-users. Some authors (e.g., Pedersen, 2016) use ‘transdisciplinarity’ in both of these senses at the same time, further adding to the confusion around definitions of terms in this area.

Furthermore, there is a noteworthy distinction to be made between interdisciplinary research, which may be produced by individuals or by teams of researchers, and interdisciplinary collaboration, which implies multiple researchers from different disciplinary backgrounds working together as a team, but which may or may not actually produce interdisciplinary research. A single researcher could engage in research activity that is informed by multiple disciplines, such as law and politics, or psychology and neuroscience. A team of researchers from different disciplinary backgrounds could work together on a project that is for all intents and purposes single-discipline research, such as when a computer scientist might provide data visualisation for a sociology project. Interdisciplinary collaboration could also refer to collaboration between researchers from different disciplinary backgrounds that does not aim to produce research outputs, such as interdisciplinary teaching (Pharo et al., 2014). To add to the conceptual untidiness in this area, the distinction between interdisciplinary research (as a noun) and interdisciplinary collaboration (as a verb) is often not made especially clear in many studies and reports (Newman, 2023).

This study focuses on one particular aspect of interdisciplinarity: the creation of teams of researchers from different disciplinary backgrounds for the purpose of working together on a joint research project – or what might be referred to as ‘interdisciplinary research collaboration’. Noting that many university strategies related to interdisciplinary research involve encouraging researchers from different disciplines to collaborate with one another (Frickel et al., 2016), and, as discussed above, that material barriers are often thought to impede this collaboration, how can interdisciplinary research collaboration be supported effectively? And more specifically, do the people who control the potential incentives and rewards for interdisciplinary collaboration agree with the people who will do the collaborative work itself on how best to support collaboration?

Method

Monash University is a large higher education and research institution located in Melbourne, Australia. It is a member of Australia’s ‘Group of Eight’ association of research-intensive universities, and at the time that the data for this study were collected, it was ranked well within the top 100 universities in the world on various global ranking systems. Monash University is a national and international research and education leader across a wide variety of academic disciplines, from medicine to architecture to sociology to astrophysics. It is competitive with other universities in Australia, and elsewhere in the world, in terms of numbers of students, numbers of academic staff, numbers of degree programs on offer, and dollar amount of external research grant funding obtained each year. Monash is a good candidate for a case study exploring material obstacles to interdisciplinary collaboration, because it has a solid supply of intellectual resources, a large and varied base of potential respondents, and significant capacity for collaboration amongst researchers from various disciplinary backgrounds. Moreover, in the Australian university sector there has been a history of grand policy statements promoting interdisciplinary research collaboration but, according to some observers, little institutional action to support researchers to collaborate across disciplines (Woelert and Millar, 2013).

In November 2021 I sent a link to an anonymous electronic survey asking about interdisciplinary research collaboration, by email, to every regular academic staff member of Monash University with a public online researcher profile. Adjunct staff, emeritus professors, and PhD students without an academic appointment were excluded, as were professional research staff (e.g., roles with titles such as ‘research officer’ or ‘lab manager’). Follow-up emails were sent in December 2021, February 2022, and March 2022. Invitations to participate were sent from a Monash University email address and contained information about the project, a link to the survey, and the primary researchers’ contact information. In all, 2,783 staff members were included in the initial invitation list. A total of 430 completed surveys were returned, for a completion rate of 16%.

The range of conceptual diversity relating to interdisciplinarity discussed above, along with the complexity of some of these definitions, could pose a problem of clarity for survey and interview respondents, who are time-poor and generally do not have the capacity to familiarise themselves with the relevant literature before answering a brief survey or participating in an interview. In order to reduce the conceptual work required by this study’s participants, in the survey and in interviews interdisciplinary research collaboration was simply defined as ‘working on a research project with partners who were affiliated with a different discipline than the one you were primarily affiliated with at that time’. All references to interdisciplinarity, interdisciplinary research, and interdisciplinary collaboration in this study should be understood in this context.

Between March and July 2021 I conducted 14 interviews with senior leaders at the university from a pool of 20 leaders identified as having roles relevant to this project. Those interviewed included 4 senior directors working in the Research and Enterprise portfolios, 7 faculty deans, 1 Deputy Vice-Chancellor, the Provost, and the Vice-Chancellor of the university. The remaining six potential interviewees were all approached but either did not respond, or declined to be interviewed, or, in the case of one individual, agreed to an interview but subsequently could not find time in their schedule to meet. Interviews were, on average, 45 minutes each; half of the interviews were conducted in person, and half remotely via video conferencing software. Thirteen interviews were audio recorded and transcribed verbatim. One interviewee declined to be recorded, and so for that interview, hand-written notes were taken. Transcripts and notes were then analysed for thematic content, yielding 11 common themes on interdisciplinarity. The Provost and the Vice Chancellor agreed to be identified (but not to have quotes attributed to them), but the other participants preferred to remain anonymous. Because the group of potential interviewees in this case study was small, direct quotes are presented here without any personal characteristics attached to them, as this may inadvertently identify the participant.

The methods described here present some obvious, but perhaps unavoidable, limitations. Like all survey-based research, the survey component of this study is vulnerable to self-selection bias: it is possible that the views of the respondents do not represent the views of the greater population under consideration, but rather only those of people who are willing to participate in a study on interdisciplinary research collaboration. However, in the sample obtained here, respondents represented a balanced variety of demographic groups, including gender, generational cohort, seniority, and career stage. In addition, survey respondents’ disciplinary backgrounds mirror the disciplinary backgrounds of the original invitation list by proportion, suggesting that the sample may have some broader representation. These details are shown in Table 1.

Table 1. Descriptive characteristics of survey respondents.

	Participants (n = 430)
	n	%
‘What gender do you identify with?’^1
woman	215	50
man	199	46
non-binary or gender diverse	3	0.7
prefer not to say	13	3
‘In what year were you born?’
pre-baby boom (<1945)	2	0.4
baby boom (1945–1964)	105	24
generation x (1965–1979)	213	50
generation y/millennial (1980–1999)	109	25
generation z (≥2000)	1	0.2
‘What is your current level of academic appointment?’^2
Level A (postdoctoral research fellow)	21	5
Level B (lecturer)	94	22
Level C (senior lecturer)	120	28
Level D (associate professor)	89	21
Level E (professor)	102	24
none of the above	4	1
‘If you have completed a PhD, in what year was your PhD conferred?’
less than 5 years ago	77	18
5–10 years ago	87	20
10–20 years ago	148	34
more than 20 years ago	101	23
‘I have not completed a PhD’	17	4
‘Irrespective of the research you are currently doing or have done in the past, which discipline is closest to the school/department/faculty/institute that you are currently primarily affiliated with?’	% of respondents (n = 430)	% of invitation list (n = 2783)
medicine/nursing/health sciences	31	33
science/maths/information technology	18	21
business or commerce	14	14
engineering	8	10
social sciences	8	4
arts and humanities	7	7
education	7	5
creative arts, design, or architecture	3	3
law	3	2

1. This is the approved phrasing of this question that is required by the Monash University Human Research Ethics Committee.

2. Lecturer and Senior Lecturer are standard academic positions with research components, equivalent to assistant professor in the North American system.

3. Percentages may not sum exactly to 100 due to rounding.

Results

Survey results

Many survey respondents claimed to believe in the traditional material obstacles to interdisciplinary collaboration. On a Likert scale ranging from 0 (strongly disagree) to 4 (strongly agree), 65% of respondents agreed or strongly agreed that interdisciplinary research requires more time and/or more resources than discipline-based research, with a mean of 2.7. On the same scale, 41% agreed or strongly agreed that it is harder to get grant funding for interdisciplinary research projects than it is to get grant funding for discipline-based research projects, with a mean of 2.2, and 44% agreed or strongly agreed that it is harder to get the results of interdisciplinary research published in good journals than it is for the results of discipline-based research, with a mean of 2.3. Clearly, respondents were more likely to be concerned about the time and resources required for interdisciplinary collaboration than they were to be concerned about grant funding or publication, where responses were far from reaching a consensus. Nonetheless, for all three of these traditional material concerns about interdisciplinary research collaboration, survey respondents were more likely to say they agree that this is a concern than to say that they disagree, with means leaning towards ‘agree’ in all three cases. These results are illustrated in Figure 1.

Figure 1. Material obstacles to interdisciplinary research collaboration: perceptions of survey respondents (n=430).

The results of these survey questions were highly internally consistent. When broken down by age cohort, gender, career stage, or academic discipline, the mean response to each question for every category was similar to the overall mean, and in every case greater than 2 (i.e., leaning towards ‘agree’), suggesting that concerns about material obstacles to interdisciplinary collaboration are independent of these personal and professional characteristics. Using the Kruskal-Wallis test for statistically significant differences between categories of data across an ordinal dependent variable, none of these relationships was found to be statistically significant at the 95% confidence level – with three exceptions: men were statistically more likely than women to report feeling that it is harder to obtain grants for interdisciplinary research and harder to publish interdisciplinary research in high-quality outlets than it is for single-discipline research, and science-based (medicine/nursing/health sciences, science/maths/information technology, and engineering) researchers were less likely than arts-based (business or commerce, social sciences, arts and humanities, education, creative arts/design/architecture, and law) researchers to report feeling that it is harder to publish interdisciplinary research in high-quality outlets than it is for single-discipline research. However, the practical difference in these cases is not immediately apparent; in all three, all of the means were over 2.0, with at least 33% of respondents in each category responding ‘agree’ or ‘strongly agree’. In other words, respondents, on average, regardless of who they are or where they come from, feel that material concerns such as time and resource requirements, grant funding, and publishing success present problems for collaborative interdisciplinary work. These results are presented in Table 2.

Table 2. Material obstacles to interdisciplinary research collaboration: perceptions of survey respondents according to personal and professional characteristics.

	Interdisciplinary research requires more time and/or more resources than discipline-based research. 0 – strongly disagree 1 – disagree 2 – neither agree nor disagree 3 – agree 4 – strongly agree			It is harder to get grant funding for interdisciplinary research projects than it is to get grant funding for discipline-based research projects. 0 – strongly disagree 1 – disagree 2 – neither agree nor disagree 3 – agree 4 – strongly agree			It is harder to get the results of interdisciplinary research published in good journals than it is for the results of discipline-based research. 0 – strongly disagree 1 – disagree 2 – neither agree nor disagree 3 – agree 4 – strongly agree
	% agree or strongly agree	mean	p-value (Kruskal-Wallis test)	% agree or strongly agree	mean	p-value (Kruskal-Wallis test)	% agree or strongly agree	mean	p-value (Kruskal-Wallis test)
overall (n=430)	65	2.7		41	2.2		44	2.3
baby boomers	56	2.6	0.055	50	2.4	0.089	39	2.2	0.970
generation x	61	2.7		38	2.2		44	2.3
millennials	77	2.8		40	2.1		47	2.3
women	60	2.6	0.116	33	2.0	0.002*	36	2.1	0.013*
men	68	2.8		50	2.4		50	2.4
arts-based	67	2.8	0.111	39	2.2	0.937	54	2.6	0.000*
science-based	61	2.6		43	2.2		36	2.1
Level A	67	2.7	0.752	39	2.2	0.420	43	2.2	0.826
Level B	65	2.7		31	2.1		41	2.3
Level C	68	2.8		41	2.2		38	2.2
Level D	63	2.7		49	2.4		51	2.4
Level E	61	2.7		46	2.2		47	2.3
No PhD	59	2.7	0.095	40	2.4	0.159	24	2.1	0.731
<5 years since PhD	68	2.6		35	2.0		47	2.3
5–10 years since PhD	66	2.7		35	2.1		38	2.2
10–20 years since PhD	68	2.9		47	2.3		44	2.2
>20 years since PhD	56	2.6		46	2.3		49	2.4

*p < 0.05.

In line with their concerns about material obstacles to interdisciplinary collaboration, survey respondents also largely favoured material solutions to these problems. When asked, ‘To what extent would any of the following make you more inclined to collaborate with researchers outside your discipline?’, with a Likert scale of 0 (no effect at all), 1 (somewhat more inclined), and 2 (much more inclined), respondents largely preferred options that involved money, time, or other employment-related rewards over less material options such as professional development or institutional change. For example, internal grant funding provided by the university for interdisciplinary collaboration was by far the most popular option, with 89% reporting that it would make them somewhat more inclined or much more inclined to collaborate, for a mean response of 1.4. The next most popular option, with 86% reporting that it would make them somewhat more inclined or much more inclined to collaborate, and also with a mean response of 1.4, were ‘other reward-based incentives’ such as funding for PhD students or for travel to academic conferences. Institutional change, such as central university coordination of interdisciplinary collaboration efforts, ranked lowest on the list, with 45% of respondents reporting it would not induce them to participate in interdisciplinary collaboration at all, for a mean response of 0.7. These results are presented in Table 3.

Table 3. Incentives to participate in interdisciplinary research collaboration, as preferred by survey respondents (n = 430) ‘to what extent would any of the following make you more inclined to collaborate with researchers outside your discipline?’.

	0 – no effect at all	1 – somewhat more inclined	2 – much more inclined	mean
More internal grants earmarked for interdisciplinary research	11%	37%	52%	1.4
Other reward-based incentives for interdisciplinary researchers, like support for PhD students or travel funding	14%	31%	55%	1.4
More cross-faculty networking opportunities	15%	41%	43%	1.3
Research performance guidelines or workload points that specifically take interdisciplinary research into account	25%	39%	36%	1.1
Teaching relief for interdisciplinary researchers	34%	29%	37%	1.0
Free professional development programs to develop interdisciplinary research and collaboration skills	36%	39%	25%	0.9
A central office that coordinates interdisciplinary research across the university	45%	39%	17%	0.7

It should be noted at this point that most Australian universities use three separate, but related, systems of performance evaluation for academic staff. First, performance standards for teaching, research, and service are published within the academic’s administrative unit (usually the school or faculty), and staff are evaluated annually according to these criteria. Staff whose performance is deemed not to meet the standards can be ‘managed’, which usually means they will be required to take on more teaching or service duties, but can also sometimes manifest as a formal reprimand. Secondly, various academic work is assigned a number of ‘workload’ points, with some tasks counting for more points than others. Work (e.g., teaching duties) is then assigned in a way that all staff members meet the required total individual workload for the year. Finally, staff members who meet specified standards for advancement may apply for promotion to the next academic level (e.g., lecturer to senior lecturer).

A closer inspection of the results presented in Table 3 reveals some more nuanced relationships. First, as might be expected, more established academics reported less interest in nearly all incentives and rewards for interdisciplinary collaboration. For travel funding or funding for PhD students, networking, workload points or performance rewards, teaching relief, professional development programs, and central university coordination of interdisciplinary research, the further away they were from the year they obtained their PhD, the less likely respondents were to report that these strategies would make them somewhat more inclined or much more inclined to collaborate across disciplines, with correlations statistically significant at the 95% confidence level. For workload points and performance, teaching relief, and professional development, the mean response for respondents with more than 20 years since obtaining their PhD crossed below 1.0, meaning that these respondents were on average more likely to report that these strategies would not induce them to participate in interdisciplinary collaboration at all. Similar results were apparent with seniority as well: more senior-ranked academics (e.g., full professors versus lecturers), regardless of when they received their PhD, were statistically less likely to report that networking, workload points or performance rewards, teaching relief, professional development programs, or central coordination of interdisciplinary research would make them more inclined to collaborate across disciplines.

Secondly, science-based researchers appeared to be motivated differently than arts-based researchers when it comes to interdisciplinary collaboration. Science-based survey respondents were statistically less likely than arts-based respondents to report that workload points or performance rewards, or teaching relief, would induce them to participate in interdisciplinary research collaboration. This may be the result of Monash University’s particular workload allocation and research performance reward system – or more specifically, the result of workload and performance accounting regimes that vary significantly across faculties at Monash University, which may not necessarily be the case at other institutions.

Thirdly, gender differences were apparent here as well. Men were statistically less likely than women to value networking, professional development, or central coordination of collaboration efforts. However, again, the practical implications of these differences are not clear. Both men and women were more likely than not to say that networking programs would persuade them to participate in interdisciplinary collaboration, and both men and women were more likely than not to say that central coordination would not encourage them to collaborate, even if the degree to which these groups agreed or disagreed with the survey questions differed. Also, there was little nominal difference and no statistical difference between male-identifying and female-identifying respondents on any of the other strategies for encouraging interdisciplinary collaboration.

And finally, older respondents were marginally more likely to value travel funding or funding for PhD students than younger respondents were, but the practical difference in this case was negligible, and no other age-related differences were statistically significant. The full results are shown in Table 4.

Table 4. Incentives to participate in interdisciplinary research collaboration: perceptions of survey respondents according to personal and professional characteristics. ‘to what extent would any of the following make you more inclined to collaborate with researchers outside your discipline?’.

	More internal grants earmarked for interdisciplinary research			Other reward-based incentives for interdisciplinary researchers, like support for PhD students or travel funding			More cross-faculty networking opportunities			Research performance guidelines or workload points that specifically take interdisciplinary research into account			Teaching relief for interdisciplinary researchers			Free professional development programs to develop interdisciplinary research and collaboration skills			A central office that coordinates interdisciplinary research across the university
	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)	% somewhat more inclined or much more inclined	mean	p-value (Kruskal-Wallis test)
overall (n=430)	89	1.4		86	1.4		84	1.3		75	1.1		66	1.0		64	0.89		56	0.72
baby boomers	94	1.5	0.421	88	1.4	0.017*	91	1.4	0.622	76	1.1	0.437	66	1.0	0.146	68	0.94	0.423	59	0.82	0.792
generation x	86	1.4		85	1.4		84	1.3		74	1.1		66	1.0		63	0.87		53	0.67
millennials	89	1.4		86	1.4		81	1.2		77	1.2		67	1.1		60	0.89		56	0.70
women	92	1.4	0.624	88	1.5	0.177	87	1.4	0.012*	75	1.1	0.843	68	1.1	0.465	75	1.07	0.000*	66	0.86	0.000*
men	87	1.4		84	1.4		82	1.2		76	1.1		66	1.0		54	0.70		45	0.57
arts-based	88	1.4	0.405	82	1.3	0.137	81	1.2	0.342	80	1.3	0.000*	73	1.2	0.002*	63	0.89	0.984	53	0.74	0.845
science-based	90	1.4		89	1.5		87	1.3		72	1.0		62	0.9		64	0.89		57	0.70
Level A	95	1.7	0.056	76	1.3	0.361	91	1.4	0.084	81	1.1	0.018*	57	0.9	0.010*	81	1.14	0.001*	52	0.67	0.001*
Level B	90	1.4		90	1.5		89	1.5		77	1.1		69	1.1		72	1.07		63	0.88
Level C	93	1.5		88	1.5		84	1.2		82	1.3		72	1.2		68	0.95		66	0.85
Level D	84	1.3		83	1.4		79	1.2		72	1.1		71	1.1		57	0.80		39	0.52
Level E	87	1.4		84	1.3		84	1.2		65	0.9		55	0.8		54	0.67		49	0.60
No PhD	94	1.4	0.163	76	1.2	0.012*	88	1.4	0.046*	94	1.5	0.003*	65	1.1	0.000*	94	1.35	0.000*	65	0.76	0.024*
<5 yrs since PhD	93	1.5		91	1.6		91	1.4		84	1.2		79	1.4		83	1.21		69	0.95
5–10 yrs since PhD	91	1.4		89	1.5		81	1.2		75	1.1		65	0.9		63	0.87		53	0.74
10–20 yrs since PhD	90	1.4		86	1.4		86	1.3		75	1.2		71	1.1		60	0.87		53	0.67
>20 yrs since PhD	83	1.3		81	1.2		82	1.1		66	0.9		53	0.8		50	0.61		49	0.58

*p < 0.05.

Interviews

University leaders interviewed for this study also discussed material obstacles to interdisciplinary research collaboration. When asked, ‘What do you think are the main barriers to interdisciplinary collaboration?’, responses included, for instance, the common concern that interdisciplinary collaboration requires more resources and effort on the part of the participating researchers than does single-discipline research. For example, according to one interviewee,

There’s an overhead on starting something new. There’s always an overhead on something new, for any individual researcher. Especially if it’s to go into a cross-disciplinary collaboration that they haven’t done before. Because they’ve got to get to know the people in the discipline and think about their own discipline differently. So there’s an overhead on it for individuals.

And according to another,

If you have to learn something new, or you have to put in more effort to go and do something other than just what you like and what you’re comfortable with, and what’s easy for you to get outputs, then you’re less likely to do it.

Interviewees were also concerned about grant funding. A common sentiment was that funding schemes incentivise single-discipline research – or, in other words, it is harder to get grant funding for interdisciplinary research than it is for single discipline research. As one respondent put it,

We recognise that interdisciplinarity is really important, but it comes at great cost. And the cost at the moment is that actually interdisciplinarity from a funding perspective is not rewarded. …if our [Australian] Research Council, and our NHMRC [National Health and Medical Research Council] are funding technically-oriented deep work, then that’s where researchers are going to go.

And, as reported by another interviewee,

Most of the Commonwealth [federal government] funding mechanisms disincentivise multidisciplinarity because assessors look to the purity of disciplinary applications and they think multidisciplinary applications diminish that intensity and purity of engagement. So multidisciplinary research is never recognised.

Interviewees also believed that it might be harder to publish the results of interdisciplinary research than the results of single-discipline research. According to one respondent, who was very concerned about the difficulties in assessing interdisciplinary research outputs for the purposes of peer review,

Interdisciplinary collaboration may or may not go into the kinds of journals that are going to be regarded as the high quality journals across a number of disciplines.

However, when it came to strategies to overcome these material barriers to interdisciplinary collaboration, university leaders appeared to have very different priorities than the academics who responded to the survey. For example, while academic survey respondents most preferred internal funding schemes to incentivise interdisciplinary collaboration, only four interviewees mentioned internal funding at all, and all four of them had negative views of this kind of incentive, arguing that it would not be effective. According to one interviewee,

I don’t think we should just throw money at it. If I hear another person talk to me about, ‘Create a big pot of money, and people can bid for that money, and that will encourage them to do it’. I don’t think you should give a single grant out for interdisciplinary work funded by the university, unless it’s got a sustainability plan beyond the university’s money.

Likewise, while workload points and performance expectations that explicitly reward interdisciplinarity ranked highly as a preferred strategy among academics, university leaders were notably antagonistic to this idea, and not a single interviewee suggested that workload points or performance criteria constitute an appropriate method of encouraging interdisciplinary collaboration. One point of view among interviewees was that changing performance criteria simply would not be effective. According to one respondent,

So why do people go into academia? Sometimes they want to just do what they want to do, they don’t like being managed… and they’re willing to trade off. So when that’s the case, you can’t change that by organisational structures. You can’t make people be motivated fundamentally differently – you can just incentivise around what their motivations are already.

And according to another, who strenuously preferred mentorship over material incentives,

No! It’s support and development. It’s not workload! Get off the metrics! No incentives, no rewards, no workload. Forget all of that. That is not how you get the best out of your researchers.

Instead, interviewees frequently cited ‘cultural change’ as the best method for encouraging interdisciplinary collaboration at the university. For example, according to one respondent,

…some of the people we’ve got in the university have been in their jobs for 20 years, and they may not yet have been fully exposed to what that change is. So there are going to be pockets of people who think, actually, ‘What’s this got to do with me?’ If you’re successful in getting funding, then you’re going to say, ‘Well why am I going to change my winning formula now?’ …It’s just cultural… And don’t get me wrong, I would say we’ve got – I can’t put a number on it – but we’ve got many many many people who naturally think in that joined up, interdisciplinary way. But they are going to come across people who have got no interest in it at all.

And according to another,

I think there’s a sense that they want to pursue their thing, and the role of the university is to reimburse them for doing what they want to do. Now, I think things have changed quite a bit, but there’s still quite a bit of residual disposition in that direction. So one of the jobs is to try and constantly remind people that we’re part of a big ecosystem of research. And it’s not ‘us and them’, or ‘you and them’ – that you are them.

This is not meant to imply, however, that interviewees thought ‘cultural change’ would be easy to implement. As one interviewee explained,

Culture is difficult. It’s the most difficult thing to master. And this is a cultural thing. It’s process, but it’s process that is going to have to impact culture. And that’s a tough one.

Discussion

In this study, both academic survey respondents and university leader interviewees valued interdisciplinary research collaboration and believed, on average, that the university should support further efforts to connect researchers across disciplines. When asked about institutional support, on a Likert scale ranging from 0 (strongly disagree) to 4 (strongly agree), 64% of survey respondents agreed or strongly agreed that interdisciplinary collaboration is encouraged by their local unit, with a mean of 2.6. On the same scale, only 21% of respondents agreed or strongly agreed that interdisciplinary collaboration is career-limiting, and 83% of respondents said they expected to collaborate with researchers outside their discipline at some point in the next three years. All 14 interviewees valued interdisciplinary collaboration, with 11 interviewees specifically linking interdisciplinary research to real-world impact.

However, academic survey respondents and university leaders did not agree on appropriate strategies for encouraging interdisciplinary research collaboration. Academics largely pointed to material incentives and rewards, such as rewards through performance criteria, internal grant schemes, and other monetary incentives, and rejected non-material solutions such as institutional change. University leaders, on the other hand, were uniformly opposed to material incentives and rewards for interdisciplinary collaboration, and tended to promote ‘cultural change’ as a preferred strategy instead.

One point of agreement between academics and university leaders on this subject was on the benefit of networking. As shown in Table 3 above, ‘More cross-faculty networking opportunities’ was the third most popular strategy among academic survey respondents, with 84% reporting that improved networking would make them somewhat more inclined or much more inclined to participate in interdisciplinary collaboration. Likewise, six out of 14 interviewees specifically mentioned networking as an effective strategy for supporting interdisciplinary collaboration, with several interviewees referring to a particular central university networking program run by the Office of the Provost in 2020 and 2021 as a successful initiative in this area. Networking may be a good compromise strategy here: it is not exactly a material incentive, in the sense that networking events do not confer time or monetary resource benefits on participants. However, networking events are material actions, requiring time and resource commitments on the part of organisers and participants, and the results – attending research presentations and actually meeting other individuals – are tangible outcomes. At the same time, networking events are highly cultural activities, resulting in exposure to new ideas and prompting participants to interact with people on a social and intellectual basis.

While average responses may have been logically consistent across survey questions, a closer look revealed significant variation along some of the personal and professional characteristics of survey respondents. A clear relationship between career stage and interest in rewards and incentives was evident in this study, with respondents at later stages of their career claiming to be less responsive to any of the strategies proposed in the survey question except for dedicated research funding. Less clear relationships were apparent for gender, age, and disciplinary background.

University academics are a diverse workforce; for example, in many places, including Australia, significant numbers of university researchers come from other countries (Scellato et al., 2015, p. 112). One might expect academics to claim a number of identities and to align with multiple social and demographic groups. The variation in survey responses discussed above suggests that different academics might respond differently to various rewards and incentives. While the precise relationships were not that clear in this limited study, there is evidence elsewhere that these relationships exist, for example in gendered experiences with research grant funding (Sato et al., 2021).

The implication, then, is that university administrators should consider more tailored strategies to support interdisciplinary research collaboration that take the needs of more specific groups of academics into consideration. Even where interdisciplinary collaboration is popular, one-size-fits-all strategies may not present the most effective option. Further research would be required to better elucidate these relationships.

In general, any major discrepancy between what academics think will most encourage them to participate in interdisciplinary collaboration, and what university leaders would like to provide, effectively constitutes its own barrier to interdisciplinary research. Academics conduct the research and do the actual collaboration, but university leaders control the resources – including material resources like time and money – that enable research in general, and interdisciplinary research in particular, to happen. In order to be effective, strategies that aspire to support or promote interdisciplinary collaboration will need to strike some kind of compromise between the needs of academics and the preferences of university leaders. This is especially pertinent in an era of perhaps increasing divisions between the perceptions and desires of the academic workforce and the needs and demands of university leadership in many countries, including Australia (Kenny, 2018).

Conclusion

The data presented here come from a single case study, and therefore any generalisation should proceed with caution. In particular, it may be that Australian universities operate in a unique policy context with respect to interdisciplinary research collaboration (see Forsyth, 2014 for an extended history of the political landscape of the Australian university sector). At least one study (Bromham et al., 2016) has shown that in applications to the Australian Research Council, Australia’s national public sector research funding organisation, interdisciplinary research projects had a lower rate of success than projects sited within a single discipline (although the difference was not tremendous – the projects deemed most interdisciplinary by Bromham et al.‘s reckoning had a success rate of about 18% versus around 24% for those deemed least interdisciplinary). Furthermore, studies have consistently shown that Australian universities tend to internalise policy priorities from government and from public sector funding organisations in their own institutional rules, which may structurally disadvantage interdisciplinary research, both in general (Gläser & Laudel, 2007; Woelert & McKenzie, 2018) and in specific interdisciplinary areas (e.g., Wright, 2022; Wright & Ville, 2017). Some have described this as a ‘paradox of interdisciplinarity’ peculiar to Australia, in which interdisciplinary research is given much lip service but is in practice not supported by institutions (Bromham et al., 2016, p. 684; Millar, 2016; Woelert and Millar, 2013). This is consistent with the findings from this study, as interview participants expressed general enthusiasm for interdisciplinary research collaboration but not for material incentives to promote it.

That being said, it is not entirely clear if, and to what extent, these dynamics are unique to Australia. Similar outcomes have been observed internationally, with universities in the United States (Espeland & Sauder, 2007), the United Kingdom (Bandola-Gill & Smith, 2022), and Italy (Abramo & D’Angelo, 2021) having been shown to internalise national-level research funding rules. Likewise, the ‘paradox of interdisciplinarity’ has been observed in Canada (Albert et al., 2015), Latin America (Vienni Baptista et al., 2019), and elsewhere. On the other hand, public policy to promote interdisciplinary research has been successfully applied in Germany (Leišytė et al., 2022). Wider comparative research might therefore be required to map out how interdisciplinary research policy diverges across various national contexts, including Australia.

In other ways, Monash University is not especially unique; many universities around the world have similar research and education profiles, similar ratios of international students, similar relationships with governments and funding agencies, and similar operating environments. As a consequence, the results presented here suggest, at the very least, that similar outcomes might be observed at other large research-intensive universities in other jurisdictions.

From one point of view, the findings of this study align with the notion of the neoliberal university, in which relationships between workers and managers (as well as those between teachers and students) are highly transactional (Ball, 2012; Canaan & Shumar, 2008; Szekeres, 2006). In such a corporatised institutional environment, it would not be surprising that workers would ask for greater material benefits and that managers, seeking to cut costs, might be reluctant to offer these benefits. It could be that if this study had asked about PhD student recruitment, or support for participation in academic conferences, or research impact, or any number of other concerns of academic university staff, the results might have been the same. On the other hand, university leaders do think that some things are worth materially incentivising, such as research projects funded by internal grant schemes, or publications rewarded through highly metricised performance regimes, among other programs. It could be that university leaders are in favour of material incentives for some activities, say, research publications, but not for others – such as interdisciplinary collaboration. Further research would be required to disentangle these issues.

In 2021–2022, after a multi-stage consultation process with the academic community, Monash University embarked on a new strategic plan that also incorporated a new university-wide research implementation strategy. Many university leaders interviewed for this project expressed confidence that the new strategic plan, and especially the consultative aspect of the plan, would resolve some of the disagreement between academics and leaders at the university on various matters related to research, including interdisciplinary research collaboration. As universities’ internal strategies continue to be re-written, and as demands on universities to demonstrate real-world impact grow, academics and leadership will need to work more closely together if they are going to advance their common aspirations towards interdisciplinary research and impact. If university leaders genuinely want to advance interdisciplinary collaboration, they will need to seek common ground with research academics or else risk continuing the paradox of interdisciplinarity.

Disclosure statement

No potential conflict of interest was reported by the author(s).