key: cord-0296556-8ze4llx7
authors: Morin, Andréanne; Helling, Britney A.; Krishnan, Seetha; Risner, Laurie E.; Walker, Nykia D.; Schwartz, Nancy B.
title: The Effect of COVID-19 on the Postdoctoral Experience: a comparison of pre-pandemic and pandemic surveys
date: 2021-11-21
journal: bioRxiv
DOI: 10.1101/2021.11.19.468693
sha: be0bf9a21315371255bb9cbc5783714a72db6742
doc_id: 296556
cord_uid: 8ze4llx7

In the interest of advocating for the postdoctoral community in the United States, we present results from survey data collected before and during the COVID-19 pandemic on the same population of postdocs. In 2019, 5,929 postdocs in the US completed a comprehensive survey, and in 2020, a subset completed a follow-up survey several months into the pandemic. The results show that the pandemic has substantially impacted postdocs’ mental health and wellness irrespective of gender, race, citizenship, or other identities. Postdocs also reported a significant impact on their career trajectories and progression, reduced confidence in achieving career goals, and negative perceptions of the job market compared to pre-COVID-19. International postdocs also reported experiencing distinct stressors due to the changes in immigration policy. Notably, having access to Postdoctoral Associations and Postdoctoral Offices positively impacted postdocs’ overall well-being and helped mitigate the personal and professional stresses and career uncertainties caused by the pandemic. Graphical Abstract Graphical Abstract of survey responses to: Why or how has your research been disrupted or not disrupted due to the pandemic? Overall, postdocs responded with feelings of loss of control as the pandemic was acting upon them and taking away their ability to complete their work.

Often unknown to those outside of the scientific community and overlooked by their awareness, but no differences in term of access to a PDA (Supplementary Figure 1F   116 and G). 117 

To directly assess the effects of COVID-19 on postdocs, we queried three general [sic] postdoc and staying home with two kids" or caring for a loved one who was/is 126 struggling with COVID-19. As one postdoc indicated, "my girlfriend has been recovering 127 from COVID-19 since March. It's a grueling process to watch and support." A large 128 number of postdocs also indicated that work progress was more difficult due to "getting 129 research done within limited shifts and hours" and an overall fear of "loss of productivity". 130 Many international postdocs were concerned about their visas and one respondent even 131 indicated that the international office at their institution told them "...you will lose your job 132 if you leave the country for any reason and are not a resident." Table 1 includes additional 133 representative responses. 134 Next, we looked at the institutional response to COVID-19, which ranged from 135 completely satisfied to completely unsatisfied (Supplementary Figure 2A) . Most 136 postdocs indicated that they were completely or mostly satisfied with their institution's 137 response to COVID-19 (59%) (Figure 2C ). In particular, postdocs with access to a Postdoctoral Affairs Office (PDO) were significantly more satisfied than those who did not 139 or were unaware of this institutional asset ( Figure 2D) . Moreover, there were no 140 differences in satisfaction to their institution's response between those with or without 141 access to a Postdoctoral Affairs Association (PDA) (Figure 2D) , or with respect to gender, 142 citizenship status, race and ethnicity, or identity (Data not shown). Notably, there was 143 also a non-negligible portion (4%) of postdocs who indicated they were completely 144 unsatisfied with their institution's response to COVID-19, with one respondent 145 commenting, "... my institution did almost NOTHING to ensure that faculty and staff can 146 be safely back at work". 147 Although the majority of postdocs indicated that all of their basic needs were met 148 during the pandemic (64%), a significant portion (36%) indicated that their needs 149 concerning mental health (21%), childcare (11%), healthcare (7%) and/or food (2%) were 150 unmet ( Figure 2E) . Additionally, 3% of postdocs wrote in responses mentioning other 151 unmet needs, including the inability to pay bills, exercise, loss of access to transportation, 152 work safety, human connections, or loss of salary, retirement benefits, or annual raise. 153 Furthermore, although the majority of postdocs indicated that all of their basic needs were 154 met, the comments indicated that the pandemic had made meeting those needs more 155 difficult; "My husband lost his job, and while we are not in danger of basic needs not being 156 met it does change some things and adds additional stress". Postdocs who had all of their 157 basic needs met were more likely to have access to a Postdoctoral Association (PDA) 158 (65% (yes (access to a PDA)) and 50% (no (no access to a PDA)); Figure 2F ). 159 Furthermore, postdocs with access to a PDO or a PDA were less likely to have their 160 mental health needs unmet (PDO: 32% (no) vs. 19% (yes); PDA: 37% (no) vs. 20% (yes), no differences were observed between those not aware and aware of a PDA or PDO at 162 their institution, Figure 2F ). Lastly, postdocs who identified as Asian (the majority of 163 whom were international (76%)) were more likely than white postdocs to report unmet 164 needs with respect to health care (12% vs. 5%) or food (5% vs. 1%) (Supplementary 165 Figure 2A ). No differences were observed according to gender, identity, or URM status 166 (data not shown).

Postdoc parents were particularly affected by pandemic-related shutdowns. While 168 we did not directly inquire of respondents in the pandemic survey whether they had 169 children (in the pre-pandemic survey, 20% of postdocs answered that they had children),

10% of respondents mentioned in comments that ensuring their children had proper care 171 was a major stressor and led to severe work disruptions. Additionally, 68% of these 172 comments were from female respondents and 32% from males suggesting a greater 173 burden of childcare for female postdocs. Overall, childcare was the 5th most frequently 174 mentioned stressor (Figure 2A-B) . Parents mentioned "I have lost childcare for my baby 175 and it has had a significant impact on my ability to write, complete research goals, and 176 apply for grants", "It was difficult to do any writing-or reading-based work because the 177 daycares were closed, and my partner and I had to divide the day into childcare/work 178 time", "Loss of productivity due to loss of childcare, feeling like I am slipping behind my 179 colleagues without children". Some reported feeling burnt out from putting in long hours 180 and mentioned lack of support from their peers and their university; "Lack of childcare 181 and intense pressure from PI to continue long hours at home", "Loss of childcare and co-182 workers not respectful of the loss of childcare", "My institution enacted strict ... "shift 183 schedules" that were outside of childcare hours so I was unable to work a full work week.

However, I was expected to produce the same (if not more) results/data to make up for 185 the time we were locked out" (more examples in Table 1) . 186 International postdocs reported more difficulty in meeting basic needs such as 187 health care (10% vs. 6%) and food (4% vs. 1%), while US citizens/PR reported more 188 difficulty in obtaining childcare (13% vs 9%) ( Figure 3A) . Additionally, international 189 respondents (n=718) expressed specific worries regarding their residency status. The Overall, 76% of respondents stated that the COVID-19 pandemic had impacted 199 their mental health, with 32% stating that it had a high or very high impact ( Figure 4A ). 200 All gender, race and ethnicity, and identity groups indicated a significant impact on mental 201 health. However, certain groups reported more of an impact than others; females and 202 third gender/non-binary reported a greater impact than males (80% and 88% vs. 68%); 203 US citizens/PR reported more of an impact than international postdocs (79% vs. 72%); 204 white and URM postdocs reported more of an impact than Asian postdocs (78% and 80% 205 vs. 68%); members of the LGBTQ community (83% vs. 75%) and postdocs with disabilities (88% vs. 76%) reported more of an impact than postdocs not identifying with 207 these groups ( Figure 4B) . 208 Parallel to this impact on mental health, access to institutional mental health 209 resources rose by 14% (Figure 4C ), which appears to be linked to an increase in Indeed, postdocs were more likely to have their mental health needs met if their institution 224 provided these resources (84%) than if their institution either did not provide them (42%) 225 or if they were unaware of these resources at their institution (68%, Figure 4E ). Access 226 to institutional mental health resources was also associated with whether postdocs had 227 their basic needs met during the pandemic. Overall, postdocs at institutions that provided 228 mental health resources were more likely to have all their basic needs met (69%) compared to those without (35%) or unaware of these resources (50%) (Figure 4F ). 230 Unsurprisingly, postdocs that did not have access to, or were unaware of mental health 231 resources at their institutions, were also more likely to have other basic needs unmet such 232 as food (8% (no), 2% (yes), 4% (not aware)) or health care (21% (no), 7% (yes), 7% (not As previously indicated ( Figure 2F ), access to a PDA and/or a PDO also increased 237 the likelihood of mental health needs being met. This trend may be due in part to a larger 238 proportion of postdocs with access to a PDO/PDA also having access to mental health 239 resources (82% and 80%) compared to those that did not (59% and 61%) or were 240 unaware (66% and 60%) (Figure 4G-H) . Furthermore, having access to a PDO/PDA was 241 associated with increased awareness of mental health resources (69% and 63% (not year earlier (see word cloud in Figure 2A -B and select comments in Table 1 ). This As previously indicated, this survey provides a unique "before-and-during" 342 opportunity to observe the effects of COVID-19 on postdoctoral life. However, there were 343 some limitations to our study. First, although the pandemic survey was conducted in a subset of the pre-pandemic respondents and therefore was more directly comparable, 345 the responses were anonymized, and we are unable to do a direct one-to-one comparison 346 of pre-pandemic to pandemic responses on an individual level. Furthermore, although we 347 were able to assess caregivers through responses to a handful of questions, including 348 the written responses, we did not directly ask if respondents were parents or caregivers, 349 limiting our ability to assess those effects more directly. Lastly, because of sample sizes, 350 we were limited in our ability to assess certain metrics for some demographics such as A. More self-identified female and third gender/non-binary and fewer self-identified male respondents completed the pandemic survey (n=1,698) compared to the pre-pandemic survey (n=5,805; Chi-squared test, p=0.0023, χ2 = 12.2). B. The majority of respondents were white in both the pre-pandemic (n=5,649) and pandemic surveys (n=1,673), with an increase in white and a decrease in Asian respondents in the pandemic survey compared to the pre-pandemic survey (Chi-squared test, p=0.0024, χ2 =12.1). C. The proportion of US citizens/PR respondents increased (Chi-squared test, p=0.0015, χ2 = 10.1; n pre-pandemic=5,813; n pandemic=1,702). D-E. As expected, the age of respondents (Chi-squared test, p=3.6x10 -14 , χ2 = 65.7; n pre-pandemic=5,825; n pandemic=1,714 ) (D) and the years of postdoc experience (Chi-squared test, p=4.3x10 -161 , χ2 = 755.8; n pre-pandemic=5,853; n pandemic=1,715) (E) both increased as we conducted the pandemic survey with a subset of the pre-pandemic respondents almost one year after the initial survey. F. The majority of respondents were in the life sciences with a statistically significant decrease in responses from those in the field of medicine in the pandemic survey (n=1,712) compared to the pre-pandemic survey (n=5,922; Chi-squared test, p=0.0012, χ2 = 32.47). PR: Permanent resident. 6, n=1,660) . Having access to a PDA significantly impacted having all their basic needs (Chi-squared test, p=0.039, χ2 = 6.5) or meeting their mental health needs (Chi-squared test, p=0.0026, χ2 = 11.9; n=1,665). disabilities also used more institutional resources (Chi-squared test, p=0.024, χ2 = 5.11) C. During the pandemic, more individuals had access to mental health resources, which was reflected in an increased awareness of these resources available at their institution (Chi-squared test, p=3.8x10 -30 , χ2 = 135.5; n pre-pandemic=5,795, n pandemic=1,713). That increase in awareness is proportional to the increase in respondents stating that their institution has available mental health resources. D. Having access (ordinal logistic regression, p= 3.54x10 -6 , OR=2.83,[95% CI:1.83-4.40]), or being aware of (ordinal logistic regression, p= 0.011, OR=1.34, [95% CI:1.07-1.67]) mental health resources reduced mental health impact during COVID -19 (n=1,710) . E. A larger portion of postdocs having access to mental health resources had their mental health basic needs met (Chi-squared test, p=2.18x10 -23 , χ2 = 104.36; n=1,722 ). F. A larger portion of postdocs having access to mental health resources had all their basic needs met (Chi-squared test, p=6.78x10 -16 , χ2 = 69.86; n=1,722). G and H. Having access to a PDO or a PDA increased the awareness (PDO (Chi-squared test, p=6.66x10 -24 , χ2 = 114.87; n=1,697); PDA (Chi-squared test, p=1.39x10 -14 , χ2 =71.01; n=1,703)) and the use (PDO (Chi-squared test, p=0.002, χ2 = 12.32; n=1,694); PDA(Chi-squared test, p=0.016, χ2 = 8.29; n=1,699)) of mental health resources. A. 14% of respondents who indicated that they are no longer a postdoc, stated that their transition was a consequence of the pandemic (n=218). B. Postdocs who transitioned due to the pandemic were more likely to be unemployed (purple) and less likely to have an academic position (red) than postdocs whose transition was not a consequence of the pandemic (Chi-squared test, p=6.69x10 -8 , χ2 = 33.04; n=205).

Uncertainty in my health, uncertainty in my partner's health, anxiety about leaving home, anxiety about how this will affect my future, depression and grievance of lost sense of "normal", lack of social interaction with others, can't visit family for forseeable [sic] future, lack of sufficient space to work from home productively, stress of fighting institutionalized racism, anxiety over changing career prospects.

Loss of morale, loss of collegial atmosphere, perception that the world is going to end, chronic anxiety about the US political situation, minority stress, worry about the health of family members, realization that working alone is terrible for my mental health, realization that nobody reads academic articles and nobody respects the professoriate, realization that the general public does not believe in science or truth.

My mental health has suffered as a consequence of being alone all the time making research more difficult….

…the extra stressors associated with the pandemic have significantly affected my mental health and ability to work effectively. …The pandemic has also taken a huge toll on my mental health which has disrupted my focus and ability to get research done.

The government released multiple rules controlling the H1-B visa of foreign workers, which make it harder for us foreigners in the job market.

1. Family getting sick and dying back home in India due to COVID-19, 2. Immigration restrictions by the government, 3. Slow pace of immigration application procedures by USCIS and US Embassies… As I am [sic] here in the US alone. My stress came from being worried about my family back in my country. and in experiencing this pandemic nearly all alone.

Having the pandemic eat into the limited amount of time I have as a postdoc here. Also being unable to travel -due to the travel ban, I cannot return home to see family (e.g. for Christmas) because I wouldn't be able to get back into the US. I was stuck in Europe for 6 months due to immigration issues (expired visa and closed embassies) and therefore was not able to do any lab work.

Relationship with PI I have been working from home, which has led to a drop in productivity. However, my PI expects me to be more productive due to "a lack of distractions." This disparity is making progress difficult….

Personally, my research has been disrupted by the constant pressure by my PI and my Institution to continue to work in lab during a pandemic. I don't feel safe working around so many people, and my complaint has been ignored by my PI and the Institution. This has caused me a lot of stress and anxiety.

… My supervisors also fell off of the map and we had almost zero contact throughout the lockdown (March -June) until we could return to the lab. Then after, the communication is still minimal and it's unclear what the status of publications are.

My PI became very micromanaging, in stark contrast to her hands-off style previously. They put a lot of pressure on me to publish and be productive during the pandemic.

Unrealistic expectations of the PI who ignored/ignores the fact that there is a pandemic and that the pandemic has an impact on research progress. First, the lab was shut down and then reopened with 25% capacity at a time.

Career/job perspectives Uncertainty/Instability in the job market as I try to find a job… Poor postdoc pay relative to the job market for my degree & experience level. … Feeling like industry/private sector is not going to be any easier to find employment in than academia with such high unemployment rates … That my project is getting behind and I will not be able to apply for grants within the window of "early career"/trainee grants.

Lack of career perspective and being unable to do my research during the final years of my postdoc.

Research Productivity I was expected to continue producing lab work while the labs were closed down! My PI encouraged me to break quarantine rules and continue work.

Lack of research output leading to fears of my career being over.

The feeling of guilt has been overwhelming. I feel like I should be doing more, but I really can't because I don't have the resources needed (e.g. mice) to do my research.

... trying to find new ways of ensuring/displaying productivity. I couldn't produce experimental results so how do represent the work that I've actually been getting done during this time. and [sic] then upon start-up, are they actually concerned and keeping student/worker safety as their primary goal.

Lockdown forced to ramp-down research to the bare minimum. Childcare restrictions have also impacted the amount of time that I can spend in the lab. Taking care of a toddler at home does not favor literature research.

An inability to balance work with childcare. My wife worked full or nearly full-time throughout the pandemic, and as a result, the bulk of childcare fell on me because I had a more flexible schedule and understanding PI. I constantly felt pressure and stress to accomplish research goals but consistently was unable to achieve anything because my children's welfare was top priority.

Lack of childcare for my school-age child. Non-COVID health concerns for my household members and paying for co-insurance and copays with the terrible insurance of my institution. My husband is unemployed and can find safe work and we are financially struggling.

Loss of productivity due to loss of childcare, feeling like I am slipping behind my colleagues without children. Lots of stress and pressure around keeping up with tasks. Unable to start any new, exciting projects that would help my career due to childcare loss.

Trying to work from home while caring for my children; it's like normal working mom guilt, but on steroids. Also, the university permanently closed the childcare center on campus (one of the best centers in the area) where our children went, so the uncertainty of being able to find quality childcare once centers reopened was exceptionally stressful. Percentage of respondents by identity. All of the identity groups were more represented in the pandemic survey compared to the pre-pandemic survey. E. Percentage of respondents by residency status, a larger percentage of respondents were US citizens and a smaller percentage of F1-OPT visa holders in the pandemic survey (Chi-squared test, χ2=36.94, p = 1.18x10 -5 ). F. Increased access to a PDO was observed during the pandemic, mainly due to an increase of awareness of such institutional resource (Chi-squared test, χ2=13.87, p = 9.73x10 -4 ). G. No differences were observed in access to a PDA before or during the pandemic. Postdocs that did not have access to mental health resources through their institutions or were unaware if their institutions had mental health resources were also more likely to have other basic needs unmet such as food (Chi-squared test, χ2=20.5, p=3.54e-5) or health care (Chi-squared test, χ2=17.7, p=1.44e-4). B. Postdocs who had access to a PDO were also more aware of mental health resources (Chi-squared test, χ2=114.18, p=6.66x10-24), C. Postdocs who had access to a PDA were also more aware of mental health resources (Chi-squared test, χ2=71.01, p=1.39e-14). 

Report of a Study Conducted Under the Auspices of the National Research Council

Employee benefits: Plight of the postdoc

Academic ecosystems must evolve to support a sustainable postdoc workforce

COVID-19 medical papers have fewer women first authors than expected

The Impact of COVID-19 on the Careers of Women in Academic Sciences, Engineering, and Medicine

Academic Productivity Differences by Gender and Child Age in Science, Technology, Engineering, Mathematics, and Medicine Faculty During the COVID-19 Pandemic

Unequal effects of the COVID-19 pandemic on scientists

American Educational Research Association et al. Voices from the field: The impact of COVID-19 on early career scholars and doctoral students

on-Early-Career-Scholars-and-Doctoral-Students

Promoting Equity for Women in Medicine -Seizing a Disruptive Opportunity

The Global Economic Impact of COVID-19

Potentially long-lasting effects of the pandemic on scientists

The impact of the COVID-19 pandemic on underrepresented early-career PhD and physician scientists

Effects of the COVID-19 pandemic on life scientists

Academic careers and the COVID-19 pandemic: Reversing the tide

Pandemic Cost NIH $16 Billion in Delayed

Colleges Have Shed a Tenth of Their Employees Since the Pandemic Began

Junior researchers hit by coronavirus-triggered hiring freezes

Uncertain prospects for postdoctoral researchers

United States National Postdoc Survey results and the interaction of gender, career choice and mentor impact

Doctors without orders

Remote Learning Barriers and Opportunities for Graduate Student and Postdoctoral Learners in Career and Professional Skill Development: A Case Study †

Impacts of COVID-19 on ecology and evolutionary biology faculty in the United States

The pandemic exposes human nature: 10 evolutionary insights

COVID-19 Disruptions Disproportionately Affect Female Academics

Postdocs need urgent financial support amid COVID conditions

The Invisible University Is COVID-19 Positive

85%) respondents did not identify any race or ethnicity and 8 (0.46%) respondents identified >=3 races or ethnicities. In the pre-pandemic survey, considering only US Postdocs, 301 (4.8%) respondents did not identify any race or ethnicity and 25 (0.4%) respondents identified >=3 races or ethnicities. * Number of individual who identify to each ethnicity/race

Postdocs who identified as Asian did not have health care (12% vs 5%, Chi-squared test, χ2=17.3, p=1.7x10 -4 ) or food (5% vs 1%, Chi-squared test

A. Females were more concerned than males (Chi-squared test, χ2=24.8, p=5.6x10-5) (n=718) and B. were more concerned about traveling (Chi-squared test, χ2=10.15, p=0.006), delays in visa renewal (Chisquared test, χ2=6.83, p=0.032) and travel bans (Chi-squared test