Routine Data for Workforce Equality Monitoring: Ethnic Inequalities in Recruitment and Workforce Representation in Nursing and Midwifery

This retrospective observational study demonstrates that routinely collected administrative data can effectively identify and monitor persistent ethnic inequalities in nursing and midwifery recruitment and workforce representation within a Scottish NHS Board, revealing that non-White staff are less likely to receive job offers and are overrepresented in lower pay bands compared to their White counterparts.

The Gist

Imagine the healthcare system as a massive, bustling hospital where nurses and midwives are the lifeblood keeping everything running. Ideally, this hospital should look like the community it serves: a diverse mix of people from all backgrounds working together at every level, from the entry-level nurses to the senior managers.

However, this study suggests that for many years, the hospital has had a "glass ceiling" and a "sticky floor" for staff from non-White backgrounds.

Here is a simple breakdown of what the researchers found, using some everyday analogies:

1. The Goal: Checking the "Employee Roster"

The researchers wanted to see if the hospital was fair. They didn't just guess; they looked at the actual "employee roster" (administrative data) from a large NHS board in Scotland. They asked two main questions:

• The Hiring Gate: When people apply for jobs, do White candidates get hired more often than non-White candidates?
• The Ladder: Once hired, do non-White staff get stuck on the bottom rungs of the ladder, while White staff climb to the top?

2. The Findings: A "Leaky Pipeline"

The study found that the system isn't working equally for everyone.

• The Hiring Gate (Recruitment): Imagine a race where everyone has to pass through a gate to get a job. The study found that for every 100 people who made it to the interview stage, White candidates were significantly more likely to be handed a job offer than non-White candidates.

  • The Analogy: It's like a relay race where the baton is passed more easily to one team than the other, even though both teams are running the same distance.
  • The Result: This gap was widest for senior roles (Band 7), where White candidates were twice as likely to get the job offer compared to non-White candidates.

• The Ladder (Workforce Representation): Once people are inside the hospital, the study looked at where they were standing.

  • The "Sticky Floor" (Band 5): This is the entry-level job. The study found that non-White staff were actually overrepresented here. It's like a waiting room that is full of non-White staff, but the door to the next room is hard to open.
  • The "Glass Ceiling" (Bands 6 & 7): As you move up to specialist and managerial roles, the number of non-White staff drops dramatically.
  • The Analogy: Imagine a pyramid. The base is wide and full of non-White staff, but as you go up to the peak (senior management), the pyramid suddenly becomes almost entirely White. The study showed this pattern hasn't changed much over the last few years.

3. The "Magic Tool": Using Existing Data

One of the coolest parts of this paper is how they did it. They didn't need to build a new, expensive survey or ask thousands of people to fill out forms.

• The Analogy: Think of the hospital's HR computer system as a giant, dusty library of records that is already being updated every day. The researchers realized they could just "read the books" that were already there to find the answers.
• Why it matters: They proved that you don't need a special project to find inequality; you just need to look at the routine data you already have, in a smart way. This is like realizing you can check your car's fuel gauge to see if you're running low, rather than needing a special mechanic to come out and measure the tank.

4. The Problem with the "Library"

The researchers also found that the "library" (the computer system) wasn't perfect.

• Sometimes, the data was missing or messy. For example, they couldn't track exactly why someone was rejected at the very first stage (shortlisting) because the system didn't always record that step clearly.
• The Analogy: It's like trying to solve a mystery, but half the clues are missing from the file. We know the suspect (inequality) is there, but we can't see exactly which step of the process is the culprit because the paperwork is incomplete.

5. The Takeaway: Why This Matters

The authors are saying: "We have the data; now we need to use it."

• Accountability: If hospitals start using these simple, routine checks every year, they can see exactly where the "leaks" in the pipeline are.
• Fairness: When people feel the system is fair, they are happier, stay longer, and provide better care.
• The Call to Action: The study suggests that Scotland (and other places without strict national rules like England's "Workforce Race Equality Standard") needs to start using these routine data checks regularly. It's not about creating new rules, but about looking at the existing numbers to make sure everyone has a fair shot at a career.

In a nutshell: The study shows that non-White nurses and midwives are facing a "double hurdle"—it's harder for them to get hired, and even harder for them to get promoted. But the good news is that we already have the tools (the routine data) to spot these problems and fix them; we just need to start using them consistently.

Technical Summary

1. Problem Statement

Health systems globally face challenges in sustaining nursing and midwifery workforces, with retention being a critical issue. Perceived inequities in recruitment and career progression are linked to lower job satisfaction, higher turnover, and reduced organizational performance. While ethnic inequalities in healthcare recruitment and senior role representation are well-documented, there is a lack of systematic, comparable data to monitor these disparities over time, particularly in Scotland. Unlike England, which utilizes the Workforce Race Equality Standard (WRES), Scotland lacks a national mechanism for routine monitoring of workforce ethnic inequalities. Existing data is often fragmented, of variable quality, and not synthesized for benchmarking or accountability.

2. Methodology

Study Design:
A retrospective observational study analyzing routinely collected administrative data from a single NHS Board in Scotland.

Data Sources & Period:

• Workforce Data: Annual "staff-in-post" data for nursing and midwifery staff from financial years 2021/22 to 2024/25.
• Recruitment Data: Pooled data on interviewed candidates and conditional job offers from 2021/22 to 2023/24 (extracted from the national JobTrain applicant-tracking system).
• Exclusions: Data for Band 8a+ was excluded due to low counts of non-White staff ($n \le 5$). Recruitment data for 2024/25 was excluded due to data quality errors.

Variables & Categorization:

• Exposure: Ethnicity, aggregated into two groups: White and Non-White (replacing the standard "Black and Minority Ethnic" label to align with the "White" category definition in the Public Sector Equality Duty, which includes White British and White minority groups).
• Outcomes:
  1. Recruitment Success: Receipt of a conditional job offer following an interview.
  2. Workforce Representation: Distribution of staff across pay bands (Bands 5, 6, and 7).
• Missing Data: Handled via a complete-case approach. Missingness was low in recruitment data (0.6%) but ranged from 3.9% to 7.6% in staff-in-post data.

Statistical Analysis:

• Recruitment Disparities: Calculated Relative Risks (RR) with 95% Confidence Intervals (CIs) to compare the likelihood of White vs. Non-White candidates receiving a conditional offer.
• Representation Analysis: Calculated Representation Quotients (RQs).
  • Formula: $RQ = \frac{\text{Proportion of subgroup in a specific band}}{\text{Proportion of subgroup in the total workforce}}$.
  • Interpretation: $RQ = 1$ (equal representation), $RQ > 1$ (overrepresentation), $RQ < 1$ (underrepresentation).
• Approach: Analyses were descriptive and unadjusted, prioritizing transparent estimates suitable for operational monitoring.

3. Key Contributions

• Novel Methodology: This is the first UK study to combine Representation Quotients (RQs) for workforce distribution with Relative Risk (RR) analysis of recruitment outcomes specifically for nursing and midwifery staff.
• Feasibility of Routine Data: Demonstrates that existing, routinely collected administrative data can generate reproducible, policy-relevant indicators for equality monitoring without the need for new data collection systems.
• Evidence of Cumulative Disadvantage: Provides empirical evidence supporting a "cumulative disadvantage" model, where small inequalities at the entry level (recruitment) compound into significant disparities at senior levels (career progression).
• Policy Gap Identification: Highlights the governance gap in Scotland compared to England and Wales, advocating for a standardized national framework similar to WRES.

4. Key Results

Recruitment Outcomes (2021/22 – 2023/24):

• Systematic Disparity: White applicants were consistently more likely to receive conditional job offers than Non-White applicants across all bands.
• Band 5 (Entry Level): White candidates had a 15% to 37% higher risk of receiving an offer compared to Non-White candidates (RR 1.15–1.37; $p < 0.001$).
• Band 6 (Specialist/Leadership): Disparities widened significantly. The RR increased from 1.29 in 2021/22 to 1.96 in 2023/24, indicating a worsening gap.
• Band 7 (Senior/Managerial): The disparity was most pronounced here, with White candidates having roughly double the likelihood of receiving an offer compared to Non-White candidates (RR ~2.10; $p < 0.005$).

Workforce Representation (2021/22 – 2024/25):

• Band 5 (Overrepresentation): Non-White staff were overrepresented at the entry level. The proportion of Non-White staff increased from 9% to 14% over the period. RQs ranged from 1.49 to 1.56.
• Bands 6 & 7 (Underrepresentation): Non-White staff were severely underrepresented in senior roles.
  • Band 6: RQs ranged from 0.48 to 0.56 (approx. 50% representation relative to the workforce).
  • Band 7: RQs ranged from 0.22 to 0.31 (approx. 25% representation relative to the workforce).
• Trend: Little change was observed in these gradients over the four-year study period, indicating persistent structural barriers.

5. Significance and Implications

For Policy and Practice:

• Accountability: The study proves that standardized indicators (RR and RQ) can be derived from routine HR data to support benchmarking and organizational accountability.
• Data Governance: It underscores the need for improved data governance, specifically regarding the consistency of recording recruitment stages (e.g., shortlisting vs. interviewing) to allow for more granular analysis of where discrimination occurs.
• Scottish Context: It provides a blueprint for Scotland to develop a national workforce equality monitoring framework similar to the English WRES, moving beyond fragmented reporting to systematic trend analysis.

Limitations:

• Single Site: Findings are from one NHS Board and may not represent the entire Scottish or UK landscape.
• Missing Data: Ethnicity data was missing for 4–8% of staff; since missingness is often non-random (higher in minority groups), the reported disparities may be conservative estimates.
• Causality: As a cross-sectional study, it cannot track individual career trajectories or prove causal mechanisms, though it strongly suggests cumulative disadvantage.

Conclusion:
The paper concludes that utilizing routine administrative data is a pragmatic and effective strategy for monitoring ethnic inequalities. Embedding these analyses into existing workforce systems can drive targeted interventions to improve retention, fairness, and the sustainability of the nursing and midwifery workforce.