A survey of working conditions within biomedical research in the United Kingdom

Survey design

A ten-question survey was designed to formally evaluate the working habits of biomedical researches. A link to the survey was included in a news article published in the Higher Education section of the Guardian newspaper on 21^st March 2016 (Riddiford, 2016a). While the primary intention was to gather information relating to UK-based biomedical scientists, the survey was also open to non-UK-based scientists from a broad range of backgrounds for comparison. The first three questions “what position are you?”, “broadly, what discipline do you work in?” and “what country do you work in?” aimed to serve as a filter to ensure the accurate analysis of UK-based biomedical scientists at different stages of their career. The following three questions “how many countries have you worked in over the past five years?”, “how old are you?” and “how long have you held this level of position?” aimed to construct a demographic census of the respondents, and to enable comparison between specific age groups. The next three questions focussed on the conditions scientists work under, asking “how many hours did you work last week?”, “how many days did you work last week?”, “what’s your annual salary in pounds sterling?”. The final question “how comfortable do you feel about your long-term prospects in research?” gave respondents the opportunity to select multiple responses, and those selecting the answer “not at all – I’m planning on leaving research” were invited to expand on their answer, and detail any factors contributing to this decision. The full list of questions and accompanying answer options are available in Supplementary File 1. The survey is still active and is hosted by Survey Monkey (https://www.surveymonkey.co.uk/r/HBP6NXX).

Data analysis

To capture as many responses as possible, data were collected between 21^st March 2016 and 6^th November 2016 (Dataset 1; Riddiford, 2017). In this time period, the survey was answered by 1,128 scientists. Initially, to give an overview of the distribution of respondents by country, IP addresses associated with each respondent were used to query a free-to-use geolocation database (http://freegeoip.net). Resulting data were then plotted on a world map, and focussed on Europe and the UK for a finer distribution of IP addresses. All the code used to generate the plots shown in Figure 1 is deposited in the Github repository https://github.com/nriddiford/Mapped-IP-addresses.

Figure 1. Distribution of responses based on geolocation of IP address.

Each point shows a single IP address and points are coloured by country. A shows IP addresses on a global map, whereas B and C show distributions across Western Europe (B) and the United Kingdom (C). Country boarders are shown as white lines in A and B.

Next, data were filtered to select only for responses from UK-based biomedical researchers (Q2 response: “biomedical sciences”; Q3: “UK”) to give a broad overview of working conditions within this cohort. Data were then further filtered to provide a career-stage-specific profile for each of the major tiers of an academic research career; the PhD, the postdoc and the principal investigator (Q1: “PhD”, “postdoc” and “principal investigator, permanent contract” or “principal investigator, non-permanent contract”). Data for each discrete profile were then analysed using a custom Perl script (Supplementary File 2) to parse downloaded data and include non-standard question answers (i.e. where respondents opted to specify a non-listed answer, or to elaborate on their selected response) in the analysis. Average salaries for each group were calculated by using the mean number of hours and salary for each cohort as queries for the web-based UK salary calculator (http://www.thesalarycalculator.co.uk/salary.php). The average salary for PhD students was calculated using the option “no NI”.

For the 299 respondents who provided a written answer to describe in detail the reasons they were planning on leaving research (Q10: “not at all – I’m planning on leaving research”), four statements were selected for each career stage as being broadly representative of the issues addressed by others in the same cohort, and are presented in Box 1–Box 3. The complete unanalysed data set for responses collected within the stated time period can be found in Dataset 1 (Riddiford, 2017; answers compromising the anonymity of respondents [IP address and personal comments] are not included).

A global view of survey respondents

A total of 1,128 responses were collected between 21^st March 2016 and 6^th November 2016, including responses from 40 countries (Figure 1 A), although the large majority were distributed across Western European countries and North America (Figure 1 A, B). Of the total responses, 667 were based in the UK (Figure 1 C).

A general profile of biomedical researchers

Of the 1,128 respondents, 900 classified themselves as biomedical scientists, 37% of whom reported having worked more than 50 hours in the week preceding the survey (12%, ≥ 60 hours). Perhaps more striking was the finding that 53% reported working more than five days the week before they answered the survey and that 15% worked every day that week (Supplementary File 1). Only 16% reported receiving an annual salary in excess of £35,000. Almost all of the respondents were PhD students or postdocs, and 98% were employed on short-term contracts.

Discrete profiles for UK-based biomedical researchers

The data were then filtered to select only for UK-based biomedical researchers, yielding a total of 520 responses, comprising 306 PhD students, 142 postdocs, 30 PIs and 42 working in roles outside of these three categories (such as research assistant), that were not included in further analyses.

PhD students. The majority of respondents to the survey were PhD students (59%), representing the youngest, and most mobile cohort, with 94% aged between 25–29 and 35% having worked in two or more countries over the past five years (Dataset 1 (Riddiford, 2017); Supplementary File 1). On average, they also reported working more hours per week than other cohorts (37% work over 50 hours a week) and the majority worked more than five days in the week before answering the survey (55%, > five days; 16%, seven days; Figure 2 - ‘PhD’). UK-based PhD students are typically funded via a tax-free stipend of between £13,000 and £20,000, which equates to an hourly salary of £6.70 (average working week reported: 48 hours; average salary: £17,000 [tax-free; excluding NI contribution]). PhD students are funded on a short-term basis, and 92% of PhD respondents have been at their current level of position for fewer than four years.

Figure 2. Graphical representation of analysed survey data.

The data are presented for three discrete career stages, the PhD (as shown in light blue box ‘PhD’), the postdoc (light purple box ‘postdoc’) and the PI (light green box ‘PI’). Venn diagrams show the distribution of respondents per cohort across age categories, with each colour representing a discrete age range (light blue: younger than 25; purple: 25–29; dark blue: 30–34; green: 35–39; yellow: 40–44; orange: 45–50; red: older than 50). Hourly salaries are shown as stylised bar charts, where the height of the bar represents hourly salary. In each case, the average salary is shown in light blue, contrasted with the minimum hourly salary shown in red. The average number of hours worked per week for each group is represented as a bar chart. In each case, the percentage of respondents per category is shown plotted against the number of hours worked per week. The number of days worked per week are represented as bubbles for each group, where bubbles are coloured by category (blue: fewer than 5 days per week; green: 5 days per week; yellow: 6 days per week; red: 7 days per week). The size of each bubble represents the relative number of respondents per group in each category. The legend shows a reference bubble (coloured grey) representing a value of 10 percent.

In response to the question “How comfortable do you feel about your long-term prospects in research?” 5% answered “comfortable”, with the vast majority expressing major concerns about one or more work-related factors. The most common reason for respondents’ lack of comfort in the prospect of a career in research was “it’s too competitive, and there aren’t enough jobs” (63%), followed by “I don’t make enough money” (45%). Surprisingly, only 28% plan on leaving academia (see Box 1 for several respondent-provided statements).

Postdoctoral researchers. The next rung on the academic ladder - and therefore the next discrete cohort analysed - is the postdoctoral research fellowship (“postdoc”), and accordingly this cohort generally comprised older respondents (65% age 30 or older; Figure 2 - ‘Postdoc’). Like PhD students, roughly a third reported having worked in two or more countries over the past five years (33%). While postdocs are also employed on a short-term basis, the number of respondents who reported being employed at the same level for four or more years was drastically higher than for PhD students (≥ four years: postdoc, 32%; PhD, 11%; ≥ ten years: postdoc, 4.5%; PhD, 0.3%), almost certainly reflecting the growing necessity of pursuing multiple postdocs on the path to becoming a full faculty member (Bourne, 2013).

Also like PhD students, postdocs work long hours - 79% reported working more than 40 hours a week, and 41% for more than five days a week. Despite their age, experience and work ethic, the average salary for biomedical postdocs in the UK is relatively low, with 75% of postdocs earning between £26,000 and £35,000 (4.5% earn more than £41,000), which constitutes an average hourly salary of approximately £14.00 (average working week reported: 45 hours; average salary: £33,000). However, despite only 7% describing themselves as “comfortable” in their long-term prospects for a career in research, only 30% plan on leaving academic research (see Box 2 for several representative reasons). The large majority that didn’t feel comfortable in a future in research and felt that they were competing for too few jobs (66% answered “It’s too competitive, and there aren’t enough jobs”) and working too hard (33% answered “I can’t keep working this hard”).

Principal investigators. The final group comprises those who identified as being a principal investigator (“PI”), and therefore represent an older and more stable cohort that PhD students or postdocs. In total, 63% of respondents in this group were employed on a permanent contract, and only 20% reported working in more than two countries over the last five years. In addition, 80% were over 35 years old and 48% reported being employed at the same level for four years or more (≥ ten years: 28%; Dataset 1 (Riddiford, 2017)). However, this category was vastly underrepresented in the survey data – only 30 individuals responded in total, and only eight were aged over 45 years – representing a major caveat in the interpretation of such data. While such low numbers are insufficient to draw any major conclusions, the data collected do provide some insight into the working habits of UK-based biomedical PIs, and particularly of younger individuals (52% employed at this level for ≤ four years). In particular, 17% in this cohort reported working over 70 hours in the week preceding the survey, and 25% worked a seven-day week (Figure 2 - ‘PI’).

Like PhD students and postdocs, the average salary from this group was relatively low (£41,000), which is particularly striking when considering the level of experience required to reach such a position, which equates to an hourly salary of just £15 (average working week reported: 49 hours; average salary: £41,000). Accordingly, a low salary was cited as a cause for concern by 38% of respondents (Q10: “I don’t make enough money”), while more respondents felt that their work/life balance was unsustainable (46%; “I can’t keep working this hard”). As in the earlier stages of a research career, roughly a third (31%) plan on leaving research for reasons such as those given in Box 3.