Executive summary
In June 2020, UK Research and Innovation (UKRI) published 5 years of harmonised data across councils (2014 to 2019) describing the diversity of funding applicants and awardees, made possible by disclosure of protected characteristics. This was followed with further releases in March 2021 of data for 2014 to 2020.
These publications indicate there is less diversity in UK environmental sciences than the Natural Environment Research Council (NERC) would wish. NERC have identified several actions to increase diversity in environmental sciences. One action is ensuring that funding processes are not resulting in differential outcomes. As such, we have undertaken a first investigation of the UKRI 2014 to 2020 dataset for NERC and provided a baseline against which interventions to improve diversity might be measured.
We drew on the UKRI data set for NERC (2014 to 2020) to consider 8 funding streams and 3 fellowship funding streams that cover the breadth of funding processes we use to assess competitive funding: an aggregated data set of 26,735 applicants.
We have explored and described differences in the diversity of applicant populations and compared this with the diversity of awardees of funding. This information is not shared with reviewers and data analysis is carried out on an anonymised basis.
We sought to understand if it was possible to identify whether there are equal outcomes for different groups across the funding processes, and if not where outcomes differ. From this we wish to generate insight as to where we may develop our processes to provide equity of opportunity.
This publication outlines our approach to this first investigation of the data and our findings. We have focused on gender and ethnicity (aggregated) as these are protected characteristics that may be inferred from a research proposal.
We have provided some understanding of diversity for the protected characteristics of age and disability (aggregated), noting additional context that our ability to explore disability across the breadth of funding opportunities is restricted by small population sizes.
Our analysis depends on disclosure of protected characteristics. Applicants and awardees who did not disclose their protected characteristic status have been excluded from the analyses, noting that this is characteristic specific.
However, only 1% (gender) and 9% (ethnicity) of the applicant pool choose not to disclose in full or part their information. Therefore, we have been able to proceed with some confidence that the understanding we have generated is relevant to the NERC population.
The population percentages described in the following are based on all applicants, which includes the small proportion of applicants who have not disclosed.
Fellowships are an important transitional career stage where consideration of the applicant is a more important part of the assessment process and so we describe separately to other funding streams.
For NERC 2014 to 2020 aggregated grants data:
- 24% of principal investigator applicants and 26 % of co-investigator applicants identified as a female
- 8% of principal investigator applicants and 12% of co-investigator applicants identified as being from an ethnic minority
- 2% of principal investigator applicants and co-investigator applicants identified as having a disability
- more applicants and awardees were in the age group 40 to 49 years than other age groups
Diversity of the applicant pool varies by funding stream and role for gender and ethnicity. Across the 8 funding streams 18% to 33% of principal investigator applicants and co-investigator applicants identified as female. Ethnic minority applications ranged from:
- 5% to 20% as principal investigator
- 4% to 38% as co-investigator
For NERC fellowship aggregated data set:
- 39% of applicants identified as female
- 9% of applicants identified as an ethnic minority
- 3% of applicants identified as having a disability
- more applicants and awardees were in the age group 30 to 39 years
For this first interrogation of our data, we used pairwise chi-squared analyses of applications and awards to assess if the outcome of self-declared populations of successful (awardees) and unsuccessful applicants was different to that which could occur by chance.
If so, we consider this likely indicates a difference in outcome for different populations of protected characteristics, and so considered if our funding processes influence this.
In describing our data sets we consider percentage award rate, defined as ‘(total number of applications with a positive decision/total number of applications) X 100’, consistent with the UKRI reporting approach.
NERC 2014 to 2020 aggregated data shows that female and ethnic minority applicants have statistically significantly lower award rates than applicants identifying as male or white, by:
- 3 percentage points (pp) and 2pp for female principal investigators and co-investigators respectively
- 6pp and 3pp for ethnic minority principal investigators and co-investigators respectively
NERC fellowship aggregated data shows that female applicants have a statistically significant higher award rate than the male applicants by 9pp and there is no significantly different outcome in award rate for white and ethnic minority applicants.
However, consideration of data in an aggregated grouping is a coarse approach to identifying if there are differences in outcome in NERC funding processes.
More granular analysis of the 8 funding streams reveals that differential outcomes are not prevalent across all funding streams.
The funding streams that showed no difference in outcome were Discovery Science, Highlight Topics, Strategic Programme Areas, and research programmes NERC delivered funded by the UKRI Strategic Priority Fund.
This list includes NERC’s 3 largest funding streams supported by NERC-core funds, which represent 71% of all applications.
NERC independent research fellowships and knowledge exchange fellowships linked to programmes also showed no differences in outcome for different protected characteristics.
The funding streams that showed differences in outcome for different protected characteristics were research investments from:
- Partnership and Opportunities
- Fund for International Collaboration
- Global Challenge Research Fund
- Newton Fund
- NERC’s open knowledge exchange fellowships
The difference in outcome did not exist in every funding stream for both female and ethnic minority candidates, nor for both principal investigators and co-investigators.
Female principal investigator award rate was not different to male award rate in 6 of the 8 funding streams. Female co-investigator award rate was not different to male award rate in 7 of the 8 funding streams. Only 1 fellowship funding stream showed a difference in outcome in favour of female applicants (open knowledge exchange fellowships).
Ethnic minority principal investigator award rate was not different to white applicant award rate in 4 of the 6 funding streams (2 of the 8 funding streams had insufficient data for statistical analysis). Ethnic minority co-investigator award rate was not different to white applicant award rate in 5 of the 8 funding streams. No fellowship funding streams exhibited a difference in outcome for ethnic applicants.
For most funding streams, we have been unable to identify funding opportunity design structures that contribute to a differential outcome. However, some opportunities in our Partnership and Opportunities funding line were atypical of most of NERC and UKRI funding opportunities, for example allowing overseas applicant to apply as principal investigators. With these opportunities excluded from the Partnership and Opportunities funding stream, there was no longer a difference in outcome between protected characteristics.
This report shares our first exploration of our data, an activity which will be undertaken in the future, benefitting also from feedback on this first report on how best to do this.
The report acts as baseline understanding to which we can return in the future. For example, considering different approaches to data analysis, changes to outcomes based on expanded data sets, the outcomes for new funding streams, and whether there is evidence the activities NERC is undertaking to support increased diversity in environmental sciences are being effective.
Such examples of activities to enhance diversity in environmental sciences include allocating funding directly to research to provide new evidence of how to improve equality, diversity and inclusion within disciplines, and trialling different approaches to ranking applications that are deemed worthy of being funded.
More detail on the actions NERC will implement can be found in its diversity and inclusion action plan.
Introduction
The 2020 UK Research and Innovation (UKRI) diversity data release revealed that diversity among those awarded funding in UK environmental sciences is limited and for some characteristics lower than the accompanying UKRI Higher Education Statistics Agency (HESA) estimates of the environmental science population (please see methods for description on the HESA data used).
For females, the principal investigator population comprises 21% of NERC award holders (2014 to 2020), which is similar to the UKRI HESA estimate that 21% of the environmental science population are female (52% of the UK population by the Office for National Statistics, 2019).
However, of the 2140 NERC grants awarded to principal investigators, only 125 (or 6%) were awarded to individuals from an ethnic minority background. This is below the UKRI estimate that 8% of the HESA environmental science population are from ethnic minorities with 14% in the UK population overall (Ethnic diversity in politics and public life).
Additionally, for NERC 2014 to 2020 aggregated data, award rates for female and ethnic minority applicants are statistically significantly lower than applicants identifying as male or white, by:
- 3 percentage points (pp) and 2pp for female principal investigators and co-investigators
- 6pp and 3pp for ethnic principal investigators and co-investigators
NERC want to understand more about the diversity of our environmental science community, to identify how we can best increase diversity in environmental sciences.
It’s important here is to ensure that our funding processes do not introduce differential outcomes. Thus, we have undertaken more detailed analysis of our 2014 to 2020 data, of 8 different funding streams and 3 fellowship categories, that cover the breadth of funding processes we use to assess competitive funding and together represent more than 90% of the applications of the UKRI data set.
This document presents the results of the consideration of NERC’s:
- aggregated data
- the individual funding streams
- the fellowship categories
It provides information about the diversity of those applying for and receiving NERC funding and whether there are differential outcomes for applicants.
We present results by the role on the funding application: principal investigator or co-investigator. UKRI standard terms and conditions only allow applications from principal investigators and co- investigators based at eligible UK research organisations.
International collaborators may be included as project partners on application forms but are not typically eligible to receive direct funding. However, there are exceptions and in some of our funding opportunities international principal investigators and co-investigators were allowed and so constitute part of the pool considered.
In the UKRI portal and for the analyses presented here, all non-UK-based principal investigators and co-investigators have been included unless stated.
Applications from those who did not disclose their protected characteristic status, on the Joint Electronic Submission (Je-S) system portal for funding applications, have been excluded from the statistical approach we used to assess if there are differences in outcome, chi-squared analyses (see ‘How we are presenting and analysing the data’).
Choosing not to declare can be characteristic specific, for example an applicant may disclose their gender but not if they have a disability. Thus, population sizes analysed vary for a given protected characteristic.
We report, as UKRI has previously, on the proportion of applicants and percentage award rate for principal investigators and co-investigators for 4 protected characteristics:
- age
- disability
- ethnicity
- gender
Percentage award rate is defined as ‘(total number of applications with a positive decision/total number of applications) X 100’, consistent with the UKRI reporting approach.
However, this report focuses on gender and ethnicity rather than age and disability, as:
- the former are more easily known or inferred from names on the grant application paperwork (and, if elicited, bias in judgement could lead to differential outcomes)
- the sample size for the latter, particularly disability, is small limiting our statistical power
- gender and ethnicity are the characteristics about which our applicant community most frequently ask for more information
Population data is calculated as the percentage of applicants for and awardees of research grant funding who self-declared protected characteristics. Approximately 1% to 9% of NERC principal investigator applicants, and 1% to 9% of co-investigator applicants do not disclose their gender or ethnicity respectively.
Applicant numbers are rounded to the nearest 5, and percentages are rounded to whole numbers.
What data we are presenting
We used the UK Research and Innovation (UKRI) harmonised data for the Natural Environment Research Council (NERC) grant awards to principal investigators and co-investigators (2014 to 2020) in our analysis, excluding the following applications (1,970 applicants):
- applications to host a postgraduate training partnership or centre, as these are large scale and involve multi-institute applications and an uncommon commissioning process
- NERC institute national capability awards which are directly commissioned from centres
- awards made to open capital funding opportunities, as the awards are often made to the institution and not an individual
- 440 applications to strategic opportunities with 100% award rates. These are collaborative, co-funded opportunities to which NERC has contributed, but that are administered by another research council or funding agency. Only successful awards directly supported by the NERC contribution are recorded in the Joint Electronic Submission (Je-S) system, from which the data is sourced. The wider application data (all the unsuccessful applications and other successful awards supported by other funders) are held by the lead council or agency. As such these data do not provide useful information regarding outcomes relative to applications for that funding opportunity and if included would skew understanding
- Global Challenge Research Fund Collective opportunities were also excluded. The panels and reviewing processes were centrally administered by UKRI. These opportunities are different to the Global Challenge Research Fund opportunities NERC directly administered for NERC applicants only which are included in the analysis
Consequently, overall numbers of principal investigator and co-investigator applicants under consideration in the analysis presented here differ from those in previous UKRI releases for NERC. We have explored separately the independent research and knowledge exchange fellowship applicants, which were included in the UKRI harmonised data set and none were excluded.
The remaining NERC grant applications in the UKRI harmonised dataset (7,555 as principal investigator and 15,660 as co-investigator), were assigned to 4 core budget funding streams (1 to 4) and to 4 National Productivity Investment Fund (NPIF) funding streams (5 to 8) led and administered by NERC. See ‘NERC core 2014 to 2020 awards’.
These funding streams, their abbreviation and the percentage of the NERC aggregated data set for each stream as principal investigator or co-investigator are as follows:
- Discovery Science: 46% and 38%
- Highlight Topics: 7% and 7%
- Strategic Programme Areas: 18% and 16%
- Partnership and Opportunities: 13% and 15%
- Newton Fund: 6% and 9%
- Strategic Priorities Fund: 5% and 5%
- Global Challenge Research Fund: 3% and 8%
- Fund for International Collaboration: 2% and 2%
Within the UKRI harmonised dataset for NERC, fellowship applications (1,350) were assigned to 3 different funding streams:
- independent research fellowships: 890, 66% of the fellowship pool
- open knowledge exchange fellowships: 215, 16% of the fellowship pool
- programme-linked knowledge exchange fellowships: 250, 18% of the fellowship pool
The funding streams were awarded as part of strategic programme investments targeted to specific areas and outcomes. The next section outlines the differences between the administration and duration of the different funding streams.
How we are presenting and analysing the data
All findings in this report are based on 2014 to 2020 (inclusive) data downloaded in March 2022 from OBIEE, the UK Research and Innovation (UKRI) grants data retrieval system.
The data were processed to remove the awards described in ‘What data we are presenting’.
This removed 1,970 applicants, leaving a sample size of 26,735 in the applicant categories of:
- principal investigators: 7,915
- co-investigators: 17,470
- fellowships: 1,350
The data were disaggregated into 8 distinct funding streams and 3 fellowship categories.
For these data sets, successful and unsuccessful applicants were summed across multiple years 2014 to 2022 and the population percentage of applicants and awardees determined for each funding stream for a given protected characteristic.
Applicants who opted not disclose their characteristic type or who did not complete the Joint Electronic Submission (Je-S) system submission were excluded from statistical analyses. This resulted in pairwise data sets for gender and ethnicity.
Application and award rates to each protected characteristic vary year on year for all the Natural Environment Research Council (NERC) funding streams. To give some indication of the level of interannual fluctuations we calculated the standard deviation of interannual variation of percentage award rate for all years for which data were available (range 2 to 6 as not all funding streams were active for 6 years).
Additionally, to test whether we could detect statistically significant differences in outcomes between different funding streams and the protected characteristics of gender and ethnicity we used SPSS Statistics to perform a pairwise chi-squared analysis comparing successful and unsuccessful outcomes at a 95% confidence level.
A significant outcome, unlikely by chance, may indicate funding processes are resulting in a differential outcome for applicants of a particular gender or ethnic characteristic. The results of these analyses are presented in tables 2, 3, and 4 in ‘Diversity analysis: applicants and awardees’.
For community diversity comparisons we use the Higher Education Statistics Agency (HESA) data estimated in the UKRI report as a proxy for the gender and ethnicity diversity profiles of the environmental science academic community eligible to apply for funding.
Some funding streams under consideration here allow for international applicants and HESA data reflects only the diversity population of the UK higher education institutes, so there are limitations to this comparison.
Process characteristics of NERC administered funding streams under consideration are outlined below.
NERC core 2014 to 2020 awards
Discovery Science
This includes:
- standard grants
- Pushing the Frontiers
- large grants
- urgency grants
Process characteristics
The key process characteristics of Discovery Science are:
- demand management applied by organisation for standard grants and Pushing the Frontiers, standard peer review and moderating panel
- large grants undergo a 2-stage assessment process: outline proposals are shortlisted by assessment panel, full stage proposals are peer-reviewed and applicants interviewed by assessment panel
- urgency grants undergo standard peer review moderated by a member of the NERC Peer Review College
Strategic Research and Innovation
This comprises:
- Highlight Topics
- Strategic Programme Areas
- Partnership and Opportunity
Process characteristics
These are directed opportunities with mixed funding approaches:
- open opportunity to Highlight Topics areas, with peer review and assessment panel. Standard peer review and panel assessments
- NERC core remit defined research topics usually peer review and subject-specific assessment panel, occasionally expert panel assessment used only
- processes may be adapted from standard NERC practice to accommodate partner process requirements. Assessment usually by peer review and subject-specific assessment panel, or straight to subject-specific expert assessment panel
Fellowships
This comprises:
- independent research fellowships
- knowledge exchange open
- knowledge exchange directed
Process characteristics
The key process characteristics of fellowships are:
- IRF proposals peer-reviewed by at least 3 reviewers, 1 of which can be an applicant-nominated reviewer. Sift panel consisting of independent experts and members of the NERC Peer Review College shortlists proposals ahead of interview. Interview questions include scientific questions raised by reviewers and non-scientific questions (such as leadership, career development or impact)
- open knowledge exchange applications go through a 2-stage process: sift panel composed of members with relevant expertise, followed by an interview with the same panel for shortlisted applicants.
- directed knowledge exchange applications are NERC core remit defined research topics, usually peer review and subject-specific assessment panel used
National Productivity Investment Funds led by NERC
Newton Fund
Awards from 2015 to 2019.
Process characteristics
Matched funding to support international partnerships. Typically, a shared peer review process between the UK and partner countries, followed by a joint subject-specific panel with up to 50% international panel members. Requirement for official development assistance (ODA) compliance.
Strategic Priorities Fund
Awards from 2018 to 2020.
Process characteristics
Multiple councils and UK partners involved in peer review and subject-specific assessment panel, occasionally expert panel assessments.
Global Challenge Research Fund
Awards from 2016 to 2020.
Process characteristics
Multiple councils and other partners involved in peer review and panels, including international panel members where appropriate. Requirement for ODA compliance.
Fund for International Collaboration
Awards from 2018 to 2020.
Process characteristics
Multiple councils and other international funding partners involved in Peer review and panels.
HESA diversity profile
The HESA diversity profile of the NERC academic community was generated by UKRI from the 2018 to 2019 staff return data by drawing on the HESA cost centres and JACS code for academic populations (staff full-person equivalent, staff (excluding atypical), academic employment function, teaching and research) that most closely reflect our remit.
There may be gaps and overlaps with other disciplines. 2018 to 2019 is the year for which UKRI have the latest available data.
A description of the data sources and limitations can be found in table 1 in the diversity results for UKRI funding data 2014 to 2019.
Considering the data and findings
When considering the data and findings please note:
- differences in research community response and purpose of funding (for example, the discipline focus of individual strategic opportunities, the amount of funding available for an opportunity) means that percentage award rate will not be the same across funding streams and within different opportunities in funding stream, and this publication therefore does not consider this
- our capacity to assess whether there are differential outcomes is dependent on disclosure of protected characteristics and we know that this information can be withheld (see top of ‘What data we are presenting’), or possibly mis-declared. For example, approximately 3% of applicants for research funding disclose that they have a disability. This is comparable to the HESA staff estimate of 3% but is lower than the Office of National Statistics (ONS) economically active population estimate of 13%. It is possible that people with disabilities are unable to choose or develop careers in environmental science and so the offset may exist, but it may also be hidden
Diversity analysis: applicants and awardees
The 2014 to 2020 aggregated data for principal investigator and co-investigator
Tables 2 to 4 show the proportion of applicants and awardees for different protected characteristics for the 2014 to 2020 aggregated grants, presented for principal and co-investigator.
In the following quantitative summaries, the corresponding Higher Education Statistics Agency (HESA) estimate of population size for the protected characteristic is given, which allows some comparison. Although a small proportion of applicants from overseas are included in the aggregated data.
Gender
In the 2014 to 2020 aggregated data, the percentage of female applicants and awardees is greater as co-investigators (26% and 28%) than principal investigators (24% and 22%). See table 2.
The HESA estimate for percentage female in the UK environmental sciences academic population is 21%.
Table 2: percentage of NERC principal investigator and co-investigator applicants that identified as female or male
| Applicant status | % female | % male |
|---|---|---|
| Principal investigator applicants | 24 | 74 |
| Principal investigator awards | 22 | 76 |
| Co-investigator applicants | 26 | 72 |
| Co-investigator awards | 28 | 70 |
Ethnicity
In the 2014 to 2020 aggregated data, the percentage of ethnic minorities as applicants and awardees is greater for co-investigators (12% and 11%) than principal investigators (8% and 6%). See table 3.
The HESA estimate for percentage ethnic minorities in the UK environmental sciences academic population is 8%.
Table 3: percentage of NERC principal investigator and co-investigator applicants that identified as being from an ethnic minority or white
| Applicant status | % ethnic minority | % white |
|---|---|---|
| Principal investigator applicants | 8 | 83 |
| Principal investigator awards | 6 | 85 |
| Co-investigator applicants | 12 | 78 |
| Co-investigator awards | 11 | 79 |
Age
In the 2014 to 2020 aggregated data, the 40–49 age group form the largest population of principal investigator and co-investigator applicants and awardees (39%). See table 4.
The HESA estimate is that this group constitutes 35% of the environmental science community. The 30 to 39 and 50 to 59 age groups are the next largest groups of applicants and awardees for UK Research and Innovation (UKRI) funding.
Table 4: percentage of NERC principal investigator and co-investigator applicants that identified as being 29 or younger, 30 to 39, 40 to 49, 50 to 59, or over 60
| Applicant status | % under 29 | % 30 to 39 | % 40 to 49 | % 50 to 59 | % over 60 |
|---|---|---|---|---|---|
| Principal investigator applicants | 1 | 26 | 40 | 27 | 7 |
| Principal investigator awards | 1 | 23 | 42 | 27 | 7 |
| Co-investigator applicants | 2 | 28 | 37 | 25 | 7 |
| Co-investigator awards | 2 | 28 | 37 | 25 | 7 |
Disability
In the 2014 to 2020 aggregated data, 2% of principal investigator and co-investigator applicants declared a disability. See table 5.
The HESA estimate for percentage staff disclosing a disability in the UK environmental sciences academic population is 4%.
Table 5: percentage of NERC principal investigator and co-investigator applicants that identified as having no disability or as having a disability
| Applicant status | % no disability | % disabled |
|---|---|---|
| Principal investigator applicants | 91 | 2 |
| Principal investigator awards | 91 | 2 |
| Co-investigator applicants | 90 | 2 |
| Co-investigator awards | 90 | 2 |
Diversity of applicants to NERC funding streams
Gender
Applications from female principal investigators comprise 21% to 27% of applications across funding streams except for:
- Global Challenge Research Fund: 33%
- Newton Fund: 18%
See table 6.
Table 6: percentage of principal investigator applicants that identified as female or male by funding stream
| Funding stream | % female | % male |
|---|---|---|
| Discovery Science | 23 | 75 |
| Highlight Topics | 21 | 77 |
| Strategic Programme Areas | 28 | 71 |
| Partnership and Opportunities | 24 | 75 |
| Fund for International Cooperation | 24 | 71 |
| Global Challenge Research Fund | 33 | 64 |
| Newton Fund | 19 | 81 |
| Strategic Priorities Fund | 27 | 71 |
Except for Newton Fund, this population is similar in size or greater than the HESA estimate of environmental sciences population that is female (21%).
The percentage of female applicants as co-investigators (18% to 33%) is greater than the HESA estimate for all funding streams except Fund for International Cooperation (18%). See table 7.
Table 7: percentage of co-investigator applicants that identified as female or male by funding stream
| Funding stream | % female | % male |
|---|---|---|
| Discovery Science | 23 | 76 |
| Highlight Topics | 26 | 72 |
| Strategic Programme Areas | 28 | 70 |
| Partnership and Opportunities | 25 | 72 |
| Fund for International Cooperation | 18 | 82 |
| Global Challenge Research Fund | 33 | 65 |
| Newton Fund | 26 | 73 |
| Strategic Priorities Fund | 28 | 70 |
Ethnicity
Applications from ethnic minority groups as principal investigators (5% to 20%) and co-investigators (4% to 38%) differ between funding streams and with investigator status. See tables 8 and 9.
Table 8: percentage of principal investigator applicants that identified as being from an ethnic minority or white by funding stream
| Funding stream | % ethnic minority | % white |
|---|---|---|
| Discovery Science | 5 | 86 |
| Highlight Topics | 5 | 86 |
| Strategic Programme Areas | 7 | 84 |
| Partnership and Opportunities | 11 | 79 |
| Fund for International Cooperation | 12 | 79 |
| Global Challenge Research Fund | 13 | 77 |
| Newton Fund | 19 | 71 |
| Strategic Priorities Fund | 9 | 83 |
Table 9: percentage of co-investigator applicants that identified as being from an ethnic minority or white by funding stream
| Funding stream | % ethnic minority | % white |
|---|---|---|
| Discovery Science | 5 | 86 |
| Highlight Topics | 4 | 87 |
| Strategic Programme Areas | 7 | 84 |
| Partnership and Opportunities | 18 | 73 |
| Fund for International Cooperation | 38 | 45 |
| Global Challenge Research Fund | 22 | 64 |
| Newton Fund | 22 | 65 |
| Strategic Priorities Fund | 11 | 82 |
For NERC core funds, except for Partnership and Opportunities, these populations are smaller than the HESA estimate of the ethnic minority pool in environmental sciences (8%):
- principal investigator
- Discovery Science: 6%
- Highlight Topics: 5%
- Strategic Programme Areas: 7%
- Partnership and Opportunity: 11%
- co-investigator
- Discovery Science: 5%
- Highlight Topics: 4%
- Strategic Programme Areas: 7%
- Partnership and Opportunity: 18%
In the collective funding streams, except for Strategic Priorities Fund, the ethnic minority application pool size as principal investigator or co-investigator is greater than the HESA estimate.
Diversity of applicants to NERC fellowships
Applications from females, ethnic minority groups and those with a disability constitute 39%, 9%, and 3% respectively.
Those aged between 30-39 years were the largest applicant group (67%), but applications were received in all age groups, with fewest applications received from those over 60 years old. See tables 10 to 13.
Table 10: percentage of NERC fellowship applicants that identified as female or male
| Applicant status | % female | % male |
|---|---|---|
| Fellowship applicants | 39 | 60 |
| Fellowship awards | 51 | 49 |
Table 11: percentage of NERC fellowship applicants that identified as being from an ethnic minority or white
| Applicant status | % ethnic minority | % white |
|---|---|---|
| Fellowship applicants | 9 | 87 |
| Fellowship awards | 6 | 90 |
Table 12: percentage of NERC fellowship applicants that identified as being as being 29 or younger, 30 to 39, 40 to 49, 50 to 59, or over 60
| Applicant status | % under 29 | % 30 to 39 | % 40 to 49 | % 50 to 59 | % over 60 |
|---|---|---|---|---|---|
| Fellowship applicants | 16 | 67 | 11 | 3 | 1 |
| Fellowship awards | 20 | 60 | 13 | 4 | 1 |
Table 13: percentage of NERC fellowship applicants that identified as having no disability or as having a disability
| Applicant status | % no disability | % disabled |
|---|---|---|
| Fellowship applicants | 90 | 3 |
| Fellowship awards | 89 | 3 |
Differentiation of fellowship applications into the different funding streams showed that applications from females (34% to 55%), ethnic minority groups (7% to 11%) and those with a declared disability (2% to 4%) differ between fellowship schemes. See tables 14 and 16.
Table 14: percentage of NERC fellowship applicants that identified as female or male by funding stream
| Applicant status | % female | % male |
|---|---|---|
| Independent research fellow | 34 | 65 |
| Open knowledge exchange fellow | 55 | 44 |
| Knowledge exchange fellowships linked with strategic programmes | 53 | 47 |
Table 16: percentage of NERC fellowship applicants that identified as being from an ethnic minority or white by funding stream
| Applicant status | % ethnic minority | % white |
|---|---|---|
| Independent research fellow | 9 | 86 |
| Open knowledge exchange fellow | 7 | 89 |
| Knowledge exchange fellowships linked with strategic programmes | 11 | 85 |
The knowledge exchange fellowships schemes received most female candidate applications, particularly the knowledge exchange open scheme.
The knowledge exchange open scheme received the fewest applications from ethnic minority candidates (7%).
Assessment of differential outcomes by funding streams
NERC 2014 to 2020 funding stream aggregated data
Gender
In the Natural Environment Research Council (NERC) 2014 to 2020 aggregated data, female principal investigators applicants have a lower award rate than male applicants by approximately 3 percentage points (pp), but a higher award rate than male applicants when applying as co-investigators.
Both outcomes were statistically significant at 95% confidence level using chi-squared analysis.
For the NERC fellowship aggregated data set, female applicants have a higher award rate than male applicants by 9pp. This outcome is significant at 95% confidence level using chi-squared analysis.
Ethnicity
In NERC 2014 to 2020 aggregated data, ethnic principal investigator and co-investigator applicants have a lower award rate than white applicants by 6pp and 3pp respectively. Both outcomes were statistically significant at 95% confidence level using chi-squared analysis.
For NERC fellowship aggregated data set, white applicants have higher award rates than ethnic minority applicants by 7pp. This outcome is not significant at 95% confidence level by chi-squared analysis.
Age
In NERC 2014 to 2020 aggregated data, different principal investigator age ranges represent between 24% and 29% of awards.
Older age groups hold more awards than the up to 29 age category (24%). For co-investigators, different age ranges represent between 27% and 29% of awards.
We did not undertake chi-squared analysis at the funding stream level as this was not possible for all age groups due to small numbers of applicants and awardees.
Disability
The award rate for principal investigators who declared a disability was 30%, and 26% for those who did not identify they had a disability. This differential outcome is statistically significant by chi-squared analysis. The numbers are too small for analysis at a funding stream level.
The award rate for co-investigators who declared a disability was 27% which is the same for co-investigators who did not declare they had a disability.
NERC award rates by funding streams
Funding stream analysis has been undertaken using data aggregated across years to provide larger data sets for analysis, hence greater statistical power, and to smooth the effect of inter-annual variation.
However, despite aggregation, some data sets are small, especially for the collective fund schemes where funding opportunity numbers have been fewer and over a shorter period.
Consideration of whether there may be a differential funding outcome is undertaken using pairwise chi-squared analysis of unsuccessful and successful outcomes for pairs of protected characteristics for gender or ethnicity (tables 18 to 21) to assess if the outcome is different to that expected by chance.
We provide the overall success and a measure of interannual variation (see ‘How we are presenting and analysing the data’).
Data sets of smaller cohorts can have larger variation in outcomes between funding opportunities and years, reflected in larger estimates of interannual variation, for example Fund for International Collaboration supported funding opportunities only where there are 2 years of data.
Gender
The pairwise chi-squared analysis of unsuccessful and successful outcomes indicates no significant difference in the award rates for female and male principal investigator or co-investigator applicants for:
- Discovery Science
- Highlight Topics
- Strategic Programme Areas
- Fund for International Collaboration
- Newton Fund
- Strategic Priorities Fund
This accounted for 79% of applicants. See table 18.
There are 2 funding streams which show differential outcomes for female principal investigator percentage award rate compared to the male applicants. Female principal investigators are less successful than male principal investigators in the Partnership and Opportunities stream. When Department for International Development opportunities are not included in Strategic Programme Areas funding streams (see end of ‘Ethnicity’ in this section for more detail), there appears a difference in outcome against female principal investigators. For co-investigators, only the Global Challenge Research Fund funding stream (8% of co-investigator applicants) shows evidence of differential gender outcome, with female CIs more successful than their male co-investigators. See table 19.
We have not been able to identify a funding process influence on these differential outcomes.
Table 18: comparison of female and male principal investigator percentage award rate by funding stream, showing with interannual variation and results of chi-square analyses for gender outcomes
| Funding stream | Female principal investigator % success rate | Female interannual variation (+/-) | Male principal investigator % success rate | Male interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Discovery Science | 19 | 5 | 19 | 6 | 0.052 | 0.8198 | 3540 | N |
| Highlight Topic | 30 | 10 | 32 | 11 | 0.234 | 0.6286 | 540 | N |
| Strategic Programme Areas | 27 | 10 | 32 | 7 | 2.255 | 0.1194 | 1350 | N |
| Strategic Programme Areas* | 26 | 10 | 32 | 11 | 4.306 | 0.0380 | 1275 | Y |
| Partnership and Opportunities | 33 | 15 | 42 | 16 | 5.307 | 0.0212 | 1023 | Y |
| Partnership and Opportunities* | 36 | 15 | 44 | 14 | 4.104 | 0.0428 | 750 | Y |
| Newton Fund | 30 | 3 | 34 | 13 | 0.359 | 0.05491 | 460 | N |
| Strategic Priorities Fund | 18 | 6 | 28 | 8 | 3.503 | 0.0613 | 340 | N |
| Global Challenge Research Fund | 27 | 4 | 19 | 0 | 1.905 | 0.1675 | 220 | N |
| Fund for International Collaboration | 26 | 25 | 25 | 20 | 0.008 | 0.9287 | 125 | N |
A significant outcome, unlikely by chance, may indicate funding processes are resulting in a differential outcome for applicants of a given protected characteristic.
Reanalysis of the Strategic Programme Areas and Partnership and Opportunities funding streams excluding Department for International Development funding opportunities is indicated by an asterisk (see end of ‘Ethnicity’ in this section).
Figures only shown where all categories (for example, outcome for a protected characteristic) have more than 4 records, and applicant numbers are rounded to the nearest 5 individuals.
Total applicants differ from UKRI harmonised dataset for NERC mentioned in ‘What data we are presenting’ as the non-disclosed and unknowns are not included in this analysis and because Global Challenge Research Fund collective fund applicants are not included.
Table 19: comparison of female and male co-investigator percentage award rate by funding stream, showing with interannual variation and results of chi-square analyses for gender outcomes
| Funding stream | Female co-investigator % success rate | Female interannual variation (+/-) | Male co-investigator % success rate | Male interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Discovery Science | 22 | 5 | 20 | 4 | 2.740 | 0.0978 | 6040 | N |
| Highlight Topic | 33 | 16 | 31 | 11 | 0.750 | 0.3866 | 1110 | N |
| Strategic Programme Areas | 31 | 11 | 33 | 8 | 0.925 | 0.3361 | 2375 | N |
| Strategic Programme Areas* | 31 | 10 | 34 | 8 | 1.805 | 0.1791 | 2205 | N |
| Partnership and Opportunities | 40 | 9 | 37 | 14 | 1.584 | 0.2081 | 2320 | N |
| Partnership and Opportunities* | 40 | 7 | 40 | 14 | 0.001 | 0.9804 | 1525 | N |
| Newton Fund | 30 | 5 | 28 | 14 | 0.593 | 0.4413 | 1425 | N |
| Strategic Priorities Fund | 26 | 3 | 29 | 6 | 0.739 | 0.000 | 740 | N |
| Global Challenge Research Fund | 30 | 5 | 28 | 14 | 17.152 | <0.0001 | 1190 | Y |
| Fund for International Collaboration | 26 | 3 | 29 | 6 | 0.012 | 0.9137 | 395 | N |
Ethnicity
There is no significant difference in the percentage award rate for ethnic minority and white principal investigators for:
- Discovery Science
- Highlight Topics
- Strategic Programme Areas
- Strategic Priorities Fund
This comprised 76% of applicants. See table 20.
However, for ethnic minority applicants, the lower percentage award rate for principal investigators in the Partnership and Opportunities and Newton Fund funding streams are unlikely by chance (19% of applications) and therefore we consider there is a difference in the outcome.
For Fund for International Collaboration and Global Challenge Research Fund, the principal investigators numbers are too small for statistical analysis (5% of applications).
There is no significant difference in the percentage award rate for ethnic minority and white co-investigators for:
- Discovery Science
- Highlight Topics
- Strategic Programme Areas
- Strategic Priority Fund
This accounted for 66% of applicants. See table 21.
However, for ethnic minority applicants, the lower percentage award rates for co-investigators in the Partnership and Opportunities, Fund for International Collaboration, Global Challenge Research Fund and Newton Fund funding streams (34% of applications) are unlikely by chance and therefore we consider there are differences in outcome for this cohort.
Table 20: comparison of ethnic minority and white principal investigator percentage award rate by funding stream, showing with interannual variation and results of chi-square analyses for ethnicity outcomes
| Funding stream | Ethnic minority principal investigator % success rate | Ethnic minority interannual variation (+/-) | White principal investigator % success rate | White interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Discovery Science | 15 | 9 | 19 | 5 | 1.897 | 0.1684 | 3280 | N |
| Highlight Topic | 25 | 25 | 27 | 7 | 0.468 | 0.4938 | 500 | N |
| Strategic Programme Areas | 25 | 13 | 30 | 8 | 1.259 | 0.2619 | 1255 | N |
| Strategic Programme Areas* | 29 | 15 | 30 | 7 | 0.088 | 0.7670 | 1190 | N |
| Partnership and Opportunities | 24 | 19 | 41 | 25 | 13.228 | 0.0003 | 940 | Y |
| Partnership and Opportunities* | 39 | 26 | 42 | 9 | 0.175 | 0.6760 | 695 | N |
| Newton Fund | 18 | 18 | 37 | 15 | 12.020 | 0.0005 | 420 | Y |
| Strategic Priorities Fund | 18 | 15 | 27 | 7 | 1.232 | 0.2672 | 325 | N |
| Global Challenge Research Fund | 7 | 4 | 25 | 5 | N/A | N/A | 205 | N/A |
| Fund for International Collaboration | 27 | 25 | 24 | 20 | N.A | N/A | 120 | N/A |
Table 21: comparison of ethnic minority and white co-investigator percentage award rate by funding stream, showing with interannual variation and results of chi-square analyses for ethnicity outcomes
| Funding stream | Ethnic minority co-investigator % success rate | Ethnic minority interannual variation (+/-) | White co- investigator % success rate | White interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Discovery Science | 17 | 8 | 20 | 4 | 1.385 | 0.2393 | 5635 | N |
| Highlight Topic | 29 | 27 | 31 | 10 | 0.145 | 0.7037 | 1025 | N |
| Strategic Programme Areas | 29 | 13 | 33 | 9 | 1.292 | 0.2556 | 2220 | N |
| Strategic Programme Areas* | 34 | 16 | 33 | 8 | 0.003 | 0.8568 | 2070 | N |
| Partnership and Opportunities | 29 | 14 | 39 | 7 | 15.377 | 0.0001 | 2145 | Y |
| Partnership and Opportunities* | 41 | 15 | 40 | 13 | 0.016 | 0.9004 | 1445 | N |
| Newton Fund | 17 | 9 | 33 | 14 | 27.903 | <0.0001 | 1260 | Y |
| Strategic Priorities Fund | 29 | 3 | 29 | 6 | 0.004 | 0.9516 | 695 | N |
| Global Challenge Research Fund | 17 | 1 | 27 | 4 | 10.435 | 0.0012 | 1050 | Y |
| Fund for International Collaboration | 5 | 18 | 14 | 46 | 7.945 | 0.0048 | 330 | Y |
Differential outcomes for ethnic minority applicants are not prevalent in funding streams that are wholly or mostly under sole NERC governance (Discovery Science, Highlight Topics, Strategic Programme Areas) but exist where there has been another partner with Partnership and Opportunities, or are part of a collective fund (Fund for International Collaboration, Global Challenge Research Fund, Newton Fund). An exception is Strategic Priorities Fund funding where the outcomes are not different from that possible by chance.
Delivery of Partnership and Opportunities, Newton Fund, and Global Challenge Research Fund investments may have often required modifications to our standard eligibility criteria, our assessment criteria, and review and assessment processes to represent all partners equitably, to accommodate partner funding constraints, or to meet the requirements of the opportunity to be met.
Such changes are partner and country dependent and retrospectively it is not likely possible to identify how this may have influenced the outcome.
However, NERC delivered 4 Official Development Assistance (ODA) compliant opportunities in the Partnership and Opportunities funding stream working in partnership with the Department for International Development (DfID).
DfID rules required us to change funding opportunity eligibility away from UK Research and Innovation (UKRI) standard policy to allow applications from principal investigators based at eligible higher education institutes (HEIs) outside of the UK.
In the DfID Partnership and Opportunities funding opportunity, this change to investigator eligibility supported an increase in ethnic minority principal investigators from 7% to 44% of the applicant pool, with 90% of these ethnic minority principal investigator applicants based outside the UK.
Reanalysing ethnic minority percentage award rate in the Partnership and Opportunities funding stream, with opportunities with ethnic minority applicants from outside the UK excluded, reduced the difference between ethnic minority and white percentage award rate from 18pp to 8pp (a difference of 10pp).
Pairwise chi-squared analysis repeated on the adjusted Partnership and Opportunities data set indicated no statistically significant difference in outcomes for ethnic minority principal investigators or co-investigators and the white applicants across the remaining Partnership and Opportunities. See table 3.
Thus, in the Partnership and Opportunities funding stream, the differential outcome overall is associated with the inclusion of overseas principal investigators.
Removing DfID investment from the Strategic Programme Areas funding stream did not change the observation of no differential outcome for a given protected ethnic characteristic.
NERC award rates by fellowships
The percentage of NERC fellowship awardees delineated by awardee identified gender and ethnicity is given in tables 15 and 17. The totals may not equal 100% due to non-disclosure.
Table 15: percentage of NERC fellowship awardees that identified as female or male by funding stream
| Awardee status | % female | % male |
|---|---|---|
| Independent research fellow | 43 | 55 |
| Open knowledge exchange fellow | 73 | 27 |
| Knowledge exchange fellowships linked with strategic programmes | 75 | 25 |
Table 17: percentage of NERC fellowship awardees that identified as being from an ethnic minority or white by funding stream
| Awardee status | % ethnic minority | % white |
|---|---|---|
| Independent research fellow | 7 | 87 |
| Open knowledge exchange fellow | 1 | 96 |
| Knowledge exchange fellowships linked with strategic programmes | 9 | 86 |
Gender
There is no significant difference in the percentage award rate for female and male applicants for independent research fellowships (IRF) or for knowledge exchange fellowships associated with research programmes (82% of applicants).
The pairwise chi-squared analysis indicated that the open knowledge exchange fellowships awards have a differential outcome in favour of female applicants. See table 21.
Ethnicity
There is no significant difference in the percentage award rate for ethnic minority and white applicants for IRF or for knowledge exchange fellowships associated with research programmes (82% of applicants). See table 22 and 23.
For the open knowledge exchange fellowship scheme, the applicant numbers are too small to undertake chi-squared analysis (18% of applications).
Table 22: comparison of gender fellowship percentage award rate by funding stream, with interannual variation and results of chi-squared analyses of outcomes
| Funding stream | Female % success rate | Female interannual variation (+/-) | Male % success rate | Male interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Independent research fellow | 12 | 10 | 8 | 11 | 22.204 | <0.0001 | 880 | N |
| Open knowledge exchange fellow | 42 | 7 | 19 | 8 | 3.542 | 0.0598 | 210 | Y |
| Knowledge exchange fellowships linked with strategic programmes | 36 | 20 | 38 | 17 | 12.756 | 0.0004 | 245 | N |
Table 23: comparison of ethnicity fellowship percentage award rate by funding stream, with interannual variation and results of chi-squared analyses of outcomes
| Funding stream | Ethnic minority % success rate | Ethnic minority interannual variation (+/-) | White % success rate | White interannual variation (+/-) | x2 | p-value | Applicant number | Significant |
|---|---|---|---|---|---|---|---|---|
| Independent research fellow | 7 | 8 | 10 | 11 | 0.612 | 0.4341 | 850 | N |
| Open knowledge exchange fellow | N/A | N/A | N/A | N/A | N/A | N/A | 205 | N/A |
| Knowledge exchange fellowships linked with strategic programmes | N/A | N/A | N/A | N/A | 0.443 | 0.5055 | 240 | N |
Chi-squared figures only shown where all categories (for example, outcome for a protected characteristic) have more than 4 records.
The applicant numbers are lower than the total numbers (paragraph 34) as not all applicants disclosed these protected characteristics.
Building on this understanding
The environmental research community
This is the first detailed analysis we have undertaken of our funding outcomes to understand if and where our processes may have differential outcomes for different protected characteristics.
This analysis has focused on ethnicity and gender and revealed that in some funding streams there is evidence of differential outcomes, sometimes in favour of applications from female individuals but not in favour of ethnic minority groups.
This publication represents the first output of a process that will be embedded in Natural Environment Research Council (NERC) activity as a responsible funder and we will seek to understand better where the processes we use or the communities we work with may be engendering a difference in outcome. For example, NERC is developing an interactive equality, diversity and inclusion dashboard that will facilitate future annual reporting and community exploration of our funding data.
NERC core funds
To improve equity in our core fund outcomes, we continuously review our standard procedures and approaches.
When improvements are identified we take positive actions to make our own processes and decision-making as equitable and transparent as possible.
Improvements already implemented include:
- undertaking equality impact assessments ahead of every new funding opportunity to draw on learning and best practice
- offering diversity, equity and inclusion training to all staff
- supporting panel members and panel chairs with resources to reinforce training from home research organisations, for example the Royal Society unconscious bias video used at panel briefings reminds people such a behavioural response is possible
- reviewing our funding opportunity texts to be as open and inclusive as possible
- introducing to the new ‘Frontiers’ offer in Discovery Science a narrative ‘Capability to Delivery’ replacement for the track record required previously
Thereafter we are seeking to find process changes that may improve equity in outcomes, for example, trialling random allocation of awards in our pilot 2022 Exploring the Frontiers opportunity to those applications considered worthy of funding.
Delivery in partnership
Understanding the causes of the significantly lower percentage award rate for ethnic minority principal investigators will require more detailed and involved analyses.
The Department for International Development (DfID) opportunities are 1 example that demonstrates how 1 operational change necessitated by working in partnership (allowing overseas principal investigators) can affect both application and outcome profiles.
To date, the increased prevalence of international principal investigators is the only mechanism we have been able to identify that clearly shapes a differential outcome. However, we will continue to explore our data to identify other points of learning and consider how we can best apply them to any future partnerships.
Future work
In this first investigation we have shown that different funding streams show differences in funding outcomes for different protected characteristics.
The next steps are to seek to understand some of the drivers of these differences and make appropriate changes to our processes where possible.
In addition to ensuring that we change processes that we understand cause a differential outcome, NERC is also seeking to increase the diversity of our applicant pool.
The demographic comparisons between NERC applicants and awardees with the NERC benchmark Higher Education Statistics Agency (HESA) data (from the UKRI portal publication) show our applicant pool to be broadly comparable to the UKRI estimate for the environmental science community:
- 8% being from an ethnic minority
- 21% being female
However, the categorisation we have is crude.
We intend to commission more detailed HESA data that will allow us to profile the potential diversity of the applicant community, from which we can make more accurate comparison and identify interventions to support applications from a more diverse group.
Although we can create opportunities, such as our recent strategic investments in enhancing diversity, equality, and inclusion in environmental sciences, it can take years to see a sustained change and we seek to work in collaboration with our research communities and research centre partners to make environmental sciences more diverse together, facilitated by our diversity, and inclusion action plan, published in June 2022.