Written evidence submitted by Wen Wang, Roger Seifert and Michael Thewall (RTR0106)

Dr Wen Wang is Association Professor in HRM DATA Analytics and Interpretation at the University of Leicester. She has a track-record on current UK workplace issues by innovatively utilising national data.  Professor Roger Seifert (Keele University) specialises in employment relations and has been working with NHS groups for over 20 years. Professor Michael Thewall (University of Wolverhampton) specialises in Cyber metrics Research. Together we make up the team exploring workplace factors which can retain Alack and Asian Minority Ethnic (BAME) staff and 50+ workers (aged between 51 and 65). Our analysis used an innovative dataset (220 NHS trust level data and its staffs’ anonymous comments from an online recruitment platform (Indeed.com). We address the question:

What are the principal factors driving staff to leave the health and social care sectors and what could be done to address them?

Our main findings:

Main findings:

• Discrimination from the public is one of the main sources causing burnout among BAME health workers and 50+ workers at work. This leads to a high ‘intention to resign’ sentiment among these groups working in NHS Trusts.
• Burnout is the decisive factor that drives 50+ health workers intention to leave the health sector for good.

What can be done to retain BAME and 50+ health workers

At the national level.

• Tough management-activated action towards discriminatory behaviour to BAME NHS front line staff
• To address staff shortage issues.

At the NHS trust level, we found that

• Positive action on health and wellbeing; fair treatment towards staff who are involved in error, near miss, or incident; and a safe culture to speak-up can reduce staff burnout and prevent resignation behaviour.
• A fair career progression is important to BAME workers
• Redeployment and flexible working conditions are important if 50+ workers are to stay.

Our findings are based on 220 NHS Trust data extracted from the NHS staff 2020. This data is matched up with 6000 anonymous staff open comments at Indeed.com.

1. Introduction

1.1 Aim of the study: We intend to explore the major causes associated with staff retention among BAME  and those aged between 50 and 65 (State Pension Age) health workers. These two groups form a substantial sector of all NHS staff, and are most likely to leave and/or express an intention to leave to find other work, to retire early, and to stop work altogether. 

1.2 Facts about BAME group and 50+ Health workers in NHS Trusts

Health workers aged 50 or over accounted for one-third of the workforce, the average retirement age among nurses is 61 years (The Lancet, 2015). One in five (21%) of whom have considered leaving the NHS for good while 54% wish to stay at their current Trust (NHS staff survey 2020). We examine those factors likely to drive them away and those good practices likely to retain them. BAME workers made up one-third of the NHS workforce, 23% of them intend to resign (ibid). Given an aging population, Brexit, and the pandemic, it is important to retain current NHS staff through understanding elements that matter most to them. For example, between January and March 2021, there were 76,500 vacancies (full-time jobs) in hospital and community services alone (Nuffield Trust, 2021), among them, 39,000 vacancies are for nurses.https://www.nuffieldtrust.org.uk/chart/the-composition-of-the-nhs-workforce-february-2020

2. Methodology

2.1 Dataset. We exacted 220 NHS trust organisational level data from the NHS staff survey (2020) to match up NHS staff’s anonymous reviews at the recruitment website Indeed.com to reveal good practices which can retain staff, especially BAME and 50+ workers.

• The NHS staff survey asks NHS staff in England about their experiences of working for their respective NHS organisations. The survey provides essential information to employers and national stakeholders about staff experience across the NHS in England. Participation is mandatory for trusts and voluntary for non-trust organisations (CCGs, CSUs, social enterprises). The survey does not cover primary care staff. Detailed national and local (organisation-level) results are also available through the links . We focused on 220 NHS trusts and use the mean (average) score of each questions among staff to capture organisational level staff collective perception of employment practices, wellbeing, safe to speak up, staff shortage etc. We used analytic weight in the analysis which is required for this type of analysis: when each observation is a mean computed from a sample size of n (number of response in each NHS trust), where n is the weight variable.

• Platform data are scraped from Indeed.com. The advancement in the internet–based technology and platforms enables us to collect hundreds of anonymous reviews, thereby detecting collective sentiment of ‘insiders’ (employees) of the same organisation.  Platform data has become an increasingly valuable data source for academics and professionals. Harvard Business School has used it regularly to capture employment patterns (Dattner, 2016) and its impact on investors and firm’s financial outcomes (Freiberg et al., 2020), more and more employees have started to use these platforms to search for jobs and for whistleblowing. For example, Indeed.com, the No.1 recruitment website has 250 million unique visitors every month and accumulated 320 million total ratings and reviews (Indeed.2021). Thousands of HR professionals have expressed concerns about lack of skills to manage these reviews and its impact on employers’ digital reputations (CIPD, 2019).   These data has been used to decode employment relations management by using real-time data (Seo&Lee, 2021; Sue et al., 2020;  Theurer et al., 2021, the Economist, 25 July 2020).

• We have scraped more than 6200 anonymous reviews by NHS staff from Indeed.com. It requires every visitor (most are job seekers) to fill in certain information before using its service to find out other people’s employment experiences of an interested employer and to post CVs on the platform. These visitors need to state their current employment status (current or former employee of an organisation reviewed), and their employment experiences including ratings and comments. Landers et al. (2019) verified the reliability of using the overall job satisfaction ratings on platform with the same period of national representative Federal Employee Viewpoint Survey of 407,789 in the United States. We thus calculated the average overall job satisfaction rating of NHS staff per organisation (between 5 and 800) to gauge their employment experiences at their respective Trust.

2.2 Research Methods

• We carried out both quantitative (in STATA 17) and qualitative analysis (Nvivo 12) within ‘the pull and push factors’ to retain BAME and 50+ health workers. The definition of main factors (variables) are included in Table 1, we checked multicollinearity (the occurrence of high inter-correlations between two factors) in Tables 2 & 3. These correlation analyses make sure these measurements make sense and also guided us for the selections of variables in the regression analysis. Analytic weight is applied to take into account the differences in number of responses across these 220 NHS trusts.
• We carried out OLS (ordinary least squares) regression (a simple and most reliable method) on percentage of staff reporting burnout, intention to resign, to retire or to stay at each of these 220 NHS trusts. The empirical results are reported in Tables 4, 5 and 6.
• The results show that our models provide strong evidence of predictive power. The model fit R2 ranges from 45% to 75%, indicates these factors included can explain up to 75% reported outcomes (burnout, intention to resign, to retire or to stay) at the NHS trust organisational level.
• When interpreting the results: positive sign (+) with three stars (***statistically significant with 99% confidence) indicates the factor in favour of the intended outcome (dependent variable) while negative sign (-) with three stars indicates the factor is against of the intended outcome.

3. Findings

3.1   Factors drive staff away

• Stress-related burn-out is the decisive factor drive NHS staff away. It was found among NHS job seekers reported at the recruitment website, Indeed.com (Figure 1). In our modelling, burnout is the main contributor, leading to 17% more likely to resign among BAME workers (Table 5). It leads to 32% more likely to retire among 50+ workers (Table 6). Those 50+ who intend to stay are those who did not burn out (Table 6).

• Factors lead to burnout are group specific. We found strong evidence that staff shortage contributing to burnout across the board (Figure 2&3) and leads 21% more likely to report burnout among 50+ workers (Table 4). However, the experience of race discrimination by patients, their relatives, and NHS staff resulted in 27% more likely among BAME workers to report burnout if one experienced such discrimination in the last 12 month (Table 4).

• Redeployment due to Covid-19 pandemic has resulted in both group reported burnout, 18% for the BAME group and 13% for the 50+ (in Table 4). Poor dealings of redeployment reduce the chance to retain staff by 15% (Table 6).   

3.2 factors which can retain staff

• Organisations with a supportive culture (Indeed rating is high), staff tend to report burn out in Table 4. However, those worked at the Covid-19 ward are more likely to stay among BAME workers in Table 5 (6% less likely to resign) and 50+ workers in Table 6 (13% more likely to stay).
• We found strong evidence to echo the recommendation made by House of Commons (June 2021): safe culture to speak up and compassionate leadership are the key to retain BAME and 50+ workers. Felt safe to raise any work-related concern is related 29% less likely to report burn out and perceive positive actions on staff health and wellbeing can reduce reported burnout by 30% among BAME workers and 24% among 50+ workers.

• Fair treatment towards staff who involved in error, or near miss an error or incident and fair career progression are the other two important factors to retain both BAME and 50+ workers. The former can reduce 18% and 8% likelihood to resign of these two groups respectively, the later can reduce 18% and 19% respectively.

4. Recommendations

4.1. Protect NHS staff from discrimination is to sustain NHS. It is important to carry on positive actions towards staff wellbeing, safe culture to speak up, fair treatment towards errors and fair career progression.  However, management endeavour to address discrimination from the public and at work is in the key to retain BAME workers.

4.2. Deploy 50+ staff appropriately can save NHS millions pounds with least investment. Redeployment was recommend to be a norm in the NHS since 2013 (Flynn and McNair, 2013), however, there still lack support system to redeploy 50+ workers at their appropriate level. For example, it is hard for a Band 7 nurse in one department to transfer to another department at the same Band. Some staff felt stuck in a rut, when redeployment occurs, it tends to be conflict laden.  There are 100,000 nurses and midwives who 50+ in the NHS staff (Nuffield Trust, 2021), if we can extend their work life to SPA (State Pension Age) instead of 61, it will save the NHS millions pounds.

To summarise, we recommend

• Tough action to protect BAME health workers from any form of discrimination at work
• Recruit and retain more staff across the board – this requires dealing with pay and terms and conditions through Review Bodies and Joint Negotiating Committees.
• Good redeployment practices to retain 50+ health workers to work longer at the NHS
• Keep improving positive action on health and wellbeing, fairness in career progression, and maintaining a safe culture within which to speak up. This should involve a less hostile attitude to trade unions and professional associations.

5. Further research

Due to the constraint of time, we could not analyse the collective sentiment of staff from platform data as we would like to, and we would like to collect more data from Glassdoor.com and NHS staff forum to identify safe culture to speak up. Since we don’t have access to NHS staff survey which is  managed by Data Access Request Service (DARS) with a lengthy scrutiny process, we could have examined the intersectional factors, such as BAME paramedics’ employment experience or 50+ midwifes’ employment experiences, thus to provide further recommendation for policymakers. This is remedial since NHS management can be pro-active in staff retention, especially if they know the issues that matter the most to specific group of staff.

References

1 Flynn M. and McNair S. Working Longer in the NHS Evidence Review: an evidence review  (2013) – commissioned by the NHS Confederation as part of the review of retirement policy in the NHS

2.Landers, R. N., Brusso, R. C., & Auer, E. M. (2019). Crowdsourcing job satisfaction data: Examining the construct validity of glassdoor. com ratings. Personnel Assessment and Decisions, 5(3), 6.

3. Nuffield Trust, 2021, the composition of the NHS workforce, February 2021

URL: https://www.nuffieldtrust.org.uk/chart/the-composition-of-the-nhs-workforce-february-2020

4. NHS staff survey, 2020 , national results. URL: https://www.nhsstaffsurveys.com/

5.House of Commons Workforce burnout and resilience in the NHS and social care https://publications.parliament.uk/pa/cm5802/cmselect/cmhealth/22/2202.htm

Figures

Figure 1 Burnout was depicted by job seekers in their comments below

Figure 2: Staff shortage and 50+ workers reported burnout by NHS trust (weighted data)

Figure 3: staff shortage and BAME workers reported burnout (weighted data)

Figure 4: Discrimination from the public and BAME workers reported burnout (weighted data)

Figure 5: Discrimination from the public and BAME workers’ intention to resign (weighted data)

Tables

Table 1 Main variables, definition and statistics

Table 2: Correlation among main variables of 50+ health workers

Note: *p<0.1; ** p<0.05; *** p<0.01

Table 3: Correlation of main variables among BAME workers

Note: *p<0.1; ** p<0.05; *** p<0.01

Table 4 OLS regression: workplace factors burn out staff (weighted data)

Note: *p<0.1; ** p<0.05; *** p<0.01. We did not include certain factors due to potential multicollinearity shown in Table 2&3

Table 5 OLS regression: workplace factors drive BAME workers to resign (weighted data)

Note: *p<0.1; ** p<0.05; *** p<0.01. We did not include certain factors due to potential multicollinearity shown in Table 2&3

Table 6: OLS regression: workplace factors push and pull 50+ health workers (weighted data)

Note: *p<0.1; ** p<0.05; *** p<0.01. We did not include certain factors due to potential multicollinearity shown in Table 2&3

Jan 2022

14