THE IMPACT OF COVID-19 ON AUSTRALIAN UNIVERSITIES

On January 25, 2020, the first case of COVID-19 was identified in Australia. Soon after, Australia closed its borders to China; by March, the border closures had effectively extended to the rest of the world. There is ongoing debate in Australia about how successfully governments managed the pandemic; however, throughout 2020 and 2021, Australia had one of the lowest COVID-19 case rates and mortality rates reported by the Organisation for Economic Co-operation and Development (Ritchie et al. Reference Ritchie, Mathieu, Rodés-Guirao, Appel, Giattino, Ortiz-Ospina and Hasell2020). Nevertheless, government mandates resulted in most universities having no on-campus lectures, research, or other activities for all or part of 2020 and 2021 (Kinash, Jones, and Crawford Reference Kinash, Jones and Crawford2021).

Due to the border closure, international students were unable to come to Australia and often deferred their study or traveled to other countries. Given the internationalization of the higher-education sector, this had a major impact on Australian universities. Before the pandemic, the sector had grown substantially to become Australia’s fourth-largest export (Australian Department of Foreign Affairs and Trade 2020). Between 2010 and 2019, income from fee-paying international students increased from 18% to 27% of total university revenue (Ferguson and Spinks Reference Ferguson and Spinks2021). In political science PhD programs, the proportion of international students among commencing cohorts doubled from approximately 20% across the years 2005–2019 (Australian Department of Education 2022). With the onset of the pandemic-related border closures, the number of people receiving a student visa decreased by approximately 33%. By September 2021, the loss of students and revenue resulted in more than 40,000 redundancies in the Australian tertiary-education sector: approximately 20% of the workforce (Kelly Reference Kelly2022).

This analysis draws on the results of an original survey that formed the basis of the authors’ broader research project on candidates’ and supervisors’ satisfaction with Australian political science PhD programs (Casey et al. Reference Casey, Rutledge-Prior, Young, Berdahl and Malloy2023). One goal of this project was to identify the level of support provided to candidates through the pandemic and the impacts that the pandemic may have had on their future career ambitions. We recognized the impact that COVID-19 has had across academic disciplines; given our own position as PhD candidates in political science at the time of the survey design, we therefore were motivated to answer van Tienoven et al.’s (Reference van Tienoven, Glorieux, Minnen, Te Braak and Spruyt2022) call for discipline-specific research on the impacts of COVID-19.

METHODS AND DEMOGRAPHICS

Our survey drew on a number of questions from two previously conducted surveys aimed at political science doctoral students (i.e., Berdahl, Malloy, and Young Reference Berdahl, Malloy and Young2020; Kefford and Morgenbesser Reference Kefford and Morgenbesser2013 Footnote ¹). It was constructed with Qualtrics and distributed during the six weeks from February 24 to April 6, 2022 (Rutledge-Prior and Casey Reference Rutledge-Prior and Casey2023).Footnote ² Survey recruitment was implemented via social media (i.e., Facebook and Twitter); the Australian Political Science Association and its working groups; and the heads of Australian political science departments, who were asked to forward the survey to their networks. The survey received responses from 109 PhD candidates and 55 PhD supervisorsFootnote ^{3 ,} Footnote ⁴ from 23 Australian universities. It asked a range of demographic questions, as well as questions on career path, doctoral training, supervision, and coursework. The questions were predominantly closed-ended, with a few allowing for free-text input (see the online appendix).

We estimate that there are no more than 300 political science academics in Australia who have supervised a PhD candidate (Australian Government 2021; Kefford and Morgenbesser Reference Kefford and Morgenbesser2013), which suggests that approximately 18% of the eligible population completed the survey. Survey respondents were broadly representative based on gender, with slight overrepresentation in the more senior levels. Because the survey included only those academics who have been on PhD supervisory panels, the resulting sample is likely to underrepresent junior academics.

More than 600 PhD candidates were enrolled in political science and policy studies in 2020 in Australia. However, this number includes candidates who were not actively studying or researching but who had maintained their enrollment. Our best estimate of active engaged PhD candidates is approximately 450. We received 109 responses—approximately 25% of the target population. PhD-candidate respondents were broadly representative relative to gender, domestic/international status, and full-/part-time status (figure 1). Australia’s “Group of Eight” (Go8) universitiesFootnote ⁵ were overrepresented: 65% compared to an estimated 45% of the population.

Figure 1 Survey Demographics

IMPACT OF COVID-19 ON THE PHD PROGRAM

We asked candidates about a range of challenges that they might have faced during the pandemic relating to academic work, research, finances, and mental health (figure 2). Of the responses, 97% indicated having experienced at least one COVID-19–related challenge and more than 70% of candidates experienced at least three challenges.Footnote ⁶

Figure 2 Candidates’ Experiences, and Supervisors’ Reports of Their Students’ Experiences, of COVID-19–Related Challenges

Pearson’s correlation coefficient = 0.932, p-value = 0.002.

The most prevalent challenge identified by candidates was reduced productivity (86%), which is consistent with Covington and Jordan’s (Reference Covington and Jordan2022) findings that 80% of PhD candidates impacted by COVID-19 experienced delays in their projects. The next two most prevalent challenges were reduced opportunities to attend conferences and other networking opportunities (74%) and mental health issues (71%). These three challenges were the same that supervisors were most likely to have identified relative to their supervisees (96%, 100%, and 93%, respectively). The agreement between the issues identified by supervisors and candidates is promising because it may indicate that some level of communication was maintained during the pandemic.Footnote ⁷

There is evidence that these pressures, including mental health issues, financial stress, and lack of future academic and nonacademic career prospects (Almasri, Read, and Vandeweerdt Reference Almasri, Read and Vandeweerdt2022; Cornwall et al. Reference Cornwall, Mayland, van der Meer, Spronken-Smith, Tustin and Blyth2019), existed before the pandemic. As one survey respondent noted, the PhD program is “already an isolating experience” that is only “aggravated by COVID.” Therefore, it is difficult to determine whether and the extent to which the impacts of these challenges have worsened relative to the pre-pandemic period. However, our findings are consistent with research in other disciplines about the impact of COVID-19 on PhD candidates (Browning et al. Reference Browning, Larson, Sharaievska, Rigolon, McAnirlin, Mullenbach and Cloutier2021; Capone, Marino, and Sang-Ah Park Reference Capone, Marino and Park2021; Covington and Jordan Reference Covington and Jordan2022; Le Vigouroux, Goncalves, and Charbonnier Reference Le Vigouroux, Goncalves and Charbonnier2021; van Tienoven et al. Reference van Tienoven, Glorieux, Minnen, Te Braak and Spruyt2022; Yassin et al. Reference Yassin, Razak, Saeed, Al-Maliki and Al-Habies2021). This consistency provides assurance about the robustness of our findings that COVID-19 has increased existing pressures on PhD candidates.

The results of our survey also suggest that PhD candidates are taking significantly longer to complete their degree relative to 2013 (figure 3)—a finding that is consistent with candidates’ reports of reduced productivity (see figure 2).

Figure 3 Year of Candidature (Full-Time Equivalent)

χ2 = 9.7518, p-value = 0.045.

A comparison of results from Kefford and Morgenbesser’s (Reference Kefford and Morgenbesser2013) survey reveals that the proportion of candidates in their fourth year or beyond has increased from 27% to 38%.Footnote ⁸ It seems likely that this delay is due in part to the impacts of COVID-19. This assumption aligns with the findings of Covington and Jordan (Reference Covington and Jordan2022), who found that approximately 40% of PhD candidates were concerned that the pandemic would delay their completion. The longer completion times also may be a result of the Australian government’s decision in response to COVID-19 to extend by six months some PhD candidates’ scholarships. Although this policy addressed issues around financial uncertainty, it is not clear whether or the extent to which it had an impact on research productivity.

DEPARTMENTAL SUPPORT FOR COVID-19–RELATED CHALLENGES

As suggested previously, both candidates and supervisors indicated a similar understanding of the challenges facing PhD candidates. However, regarding the level of departmental support for candidates, we found a significant difference in their perceptions. The survey indicates that whereas there is broad agreement across the two groups that some level of support was provided, supervisors reported that more support had been provided. Figure 4 shows that although almost 20% of candidates reported having been offered no support by their department,Footnote ⁹ less than 5% of supervisors believed this was the case. Likewise, 91% of supervisors reported that candidates were provided with “some” or “a lot” of support compared to 73% of candidates. This disconnect is cause for concern. It may mean that although there was broad agreement about the issues that candidates faced, there was insufficient discussion between candidates and supervisors about whether there was (appropriate) support in place to manage these challenges.

Figure 4 Departmental Support for Candidates Relative to COVID-19 Challenges

Fisher’s p-value = 0.034.

When we considered levels of satisfaction with departmental support relative to COVID-19, candidates were more likely than supervisors to report being “not at all satisfied” with the level of support that was offered (30% and 20%, respectively). However, the difference in satisfaction levels was not statistically significant (Fisher’s p-value = 0.355),Footnote ¹⁰ which suggests that both supervisors and candidates are in some agreement that there is more that departments can do to assist candidates.

As discussed previously, we may question how much the pandemic, in fact, impacted these results or whether the results reflect dissatisfaction with levels of departmental support in general. Exploring these results, we found that candidates’ views on whether they are being provided sufficient support from their supervisors and other staff does not appear to impact their satisfaction with COVID-19–related support (Fisher’s p-value = 0.150). That is, there is no difference in satisfaction levels across those who believe they are receiving enough support in general and those who do not: substantial proportions of both groups are dissatisfied with the COVID-19–related support offered.

We can also discern how the satisfaction with departmental support might have changed due to the pandemic by comparing satisfaction levels before and after 2020. Our comparison of identical questions on Kefford and Morgenbesser’s (Reference Kefford and Morgenbesser2013) survey with our own suggests that dissatisfaction levels indeed changed across the pre- and inter-pandemic periods. We found a statistically significant decrease in the number of PhD candidates in 2013 compared to 2022 who reported that they were receiving sufficient support from their supervisors and/or their department (figure 5). Although this measure cannot definitively explain what is driving this change, it nevertheless lends credence to the notion that levels of satisfaction with support may have been impacted by the responses of departments to the pandemic.

Figure 5 Satisfaction with Departmental Support Over Time

χ2 = 6.1993, p-value = 0.045.

IMPACT OF COVID-19 ON FUTURE CAREER PLANS

Our results indicate a marked decrease in the proportion of candidates in 2022 who were planning a career in academia relative to those with the same plans in 2013 (figure 6).Footnote ¹¹ Whereas a majority of 2022 candidates (54%) indicated that the pandemic did not have an impact on their plans for employment after completing their PhD,Footnote ¹² as shown in figure 2, more than half of candidates (56%) indicated that they were concerned about their future career prospects in academia as a result of the pandemic. This broadly reflects the findings in the Covington and Jordan (Reference Covington and Jordan2022) study, in which 68% of PhD students surveyed indicated that they were concerned the pandemic would have a negative impact on their academic career.

Figure 6 Plans to Pursue an Academic Career (2013 Versus 2022)

There is also some evidence to suggest that PhD candidates who started their PhD program since 2020 are more likely to report that COVID-19 had “no impact” on their career plans relative to those who started before the onset of the pandemic (figure 7). This may indicate that the former quickly adjusted their expectations to the “new normal.” Nevertheless, even among this cohort, 33% indicated that the pandemic has had some degree of impact on their employment plans.

Figure 7 Impact of COVID-19 on Post-PhD Career Plans, by Start of PhD Program

χ2 = 4.9145, p-value = 0.086.

We also found a substantial gap between the views of PhD candidates about their future career plans and those of their supervisors relative to their candidates’ career plans. Candidates are much less likely to indicate that their plans changed due to the pandemic than their supervisors (figure 8).

Figure 8 Views on Whether COVID-19 Has Changed Post-PhD Career Plans

Fisher’s p-value = 0.003.

This potential disconnect between candidates and supervisors is perhaps concerning if it indicates that there has been insufficient communication between candidates and supervisors about the impact COVID-19 has had on career objectives. These results are perhaps not surprising, however, when we consider a further disconnect between the views of candidates and supervisors relative to the latter’s career-mentoring style (figure 9).

Figure 9 Perceptions of Supervisors’ Career-Mentoring Style

Fisher’s p-value = 1.543e-12.

Most candidates (56%) suggested that their supervisors did not provide any career mentoring, compared to 2% of supervisors. This perceived lack of career mentoring was also evident in the free-text responses. One candidate commented that supervisors needed to “listen to the needs of their PhD students rather than assuming the students’ desired career paths” and to “listen more to PhD candidate’s particular interests as they relate to careers.”