Nursing Compensation Nursing Salary

Hospital staff shortage: the role of the competitiveness of pay of different groups of nursing staff on staff shortage Jean-Baptiste Combes a, Robert Francis Elliottb and Diane Skåtunb

aUniv Rennes, EHESP, CNRS, ARENES – UMR 6051, Rennes, France; bHealth Economics Research Unit, University of Aberdeen, Aberdeen, Scotland

ABSTRACT Shortages of nursing staff in OECD countries have been a preoccupation for policy makers. Shortages of staff may be the consequence of uncompetitive pay. In the private sector, employ- ers in different regions can offer different pay rates to reflect local amenities and cost of living. Hospitals in the UK however cannot set the pay for their employees, and as a result they might therefore incur staff shortages. Moreover, occupational groups do not operate in isolation. Shortages of staff may also be the consequence of the competitiveness of pay of an alternative group of staff. This is investigated using two distinct groups of nursing staff: assistant nurses (ANs) and registered nurses (RNs) working in English hospitals in 2003–2005 using national-level data sets. We find that an increase by 10% of the pay competitiveness of RNs decreases the shortage of both the RNs and of ANs by 0.6% and 0.4%, respectively.

KEYWORDS Wage regulation; local pay; standardized spatial wage differentials; shortage of staff; nurses; pay competitiveness; labour substitution

JEL CLASSIFICATION I18; J31; I11

I. Introduction

Shortage of nursing staff in OECD countries is a key policy issue. This article shows that where the competitiveness of pay increases, shortage of nur- sing staff is reduced. In particular, we consider the interaction of pay competitiveness between two groups of nurses. Previous literature in this area has considered single occupations in isolation. However, the health service is characterized by different occupations working as members of a larger team. Where the pay competitiveness of one group has an impact on their own shortage, it may also impact on another group in the pre- sence of underlying links between the two. We examine this using two distinct staff groups within nursing: registered nurses (RNs) who are qualified to a degree level and licenced to practise as such under the UK nurse regulatory authority and assistant nurses (ANs) who have no formal nur- sing qualification and operate out with the regu- latory authority.

This extends the literature which has already established the link between the local competitive- ness of pay of RNs with local variations in vacancy rates (Elliott et al. 2007, 2010). We show in this

article that there is no evidence that hospitals have difficulties in hiring ANs because of their own pay competitiveness, but an increase of 10% of RN pay competitiveness reduces the AN vacancy rates by 0.4% and RN vacancy rates by 0.6%. Where pay for RNs is uncompetitive, ANs might foresee higher workloads in the current post or uncompe- titive pay in potential future roles if they were to move up the career path to become RNs. Either way this would work as deterrent.

II. Pay competitiveness

Pay in the private sector is expected to differ between geographical areas within the UK due to differences in gender, age, education, industrial and occupational composition of the workforce. Higher pay in some areas of the country is also expected where the cost of living is higher while higher pay is also necessary to compensate for less pleasant working environments or amenities (Smith 1776; Rosen 1986). It is this latter form of geographical variation in pay that captures the competitiveness of pay through its ability to adjust to local labour market conditions. In the UK, we

CONTACT Jean-Baptiste Combes [email protected] Lecturer in Health Economics at the School of Public Health, 2 Avenue Gaston Berger, 35043 Rennes cedex, France.

APPLIED ECONOMICS 2018, VOL. 50, NO. 60, 6547–6552 https://doi.org/10.1080/00036846.2018.1490000

© 2018 Informa UK Limited, trading as Taylor & Francis Group

expect this to feature within the private sector where only a small proportion of staff have their pay set by collective bargaining and an even lower proportion have their pay set nationwide (Metcalf, Hansen, and Charlwood 2001). Thus, pay in the private sector is free to adjust to reflect local labour market conditions.

Pay in the public sector in the UK is set nationally by quasi arbitration bodies. Pay set in this way or by collective bargaining is less likely to vary locally (Traxler and Brandl 2011). In the UK, where the healthcare sector is dominated by public sector pro- vision within a National Health Service (NHS), nurses pay is set nationally by the NHS Pay Review Body. Therefore, pay offered by hospitals will not adjust to the local cost of living or working environ- ment conditions (Elliott et al. 2007).Where local pay is not able to reflect local labour market conditions, hospitals will have difficulties in attracting staff.

The methodology in our article follows that as outlined by Elliott et al. (2006), (2010). Using the private sector as a benchmark, we extract out the pure spatial element of wage variation by standar- dizing for other elements known to affect pay. This generates a standardized spatial wage differential (SSWD). Thus, assuming the private sector repre- sents the equilibrium in the local labour market: it is the standardized pay that is required to attract an employee to this geographical area. Given the national wage structure within the public sector, we expect less variation in standardized wage dif- ferentials within the public sector. By comparing the patterns in SSWDs between the private and public sector, we can construct a measure of the local pay competitiveness of the public sector: the public–private pay gap.

We identify appropriate private sector com- parators for both ANs and RNs using the Standard Occupational Classification (SOC) sys- tem, where individuals are classified according to employment characteristics (see Office National Statistics (2000)). We use SOC groups represent- ing ‘personal services occupations’, and ‘associate professional and technical occupations’ as the pri- vate sector comparators for ANs and RNs, respec- tively. The competitiveness of pay of ANs and RNs in any area can therefore be defined as the gap between their SSWD and that of their comparator within the private sector:

gaps ¼ SSWDs PUBLIC � SSWDc

PRIVATE (1)

where s = (AN, RN) and c = (SOC6, SOC3) These two public–private pay gap measures for

ANs and RNs are then used to consider the effect of pay competitiveness on both their own shortages but also to test for interactions of pay competitive- ness on the measure of shortage of the other.

III. Data

Our empirical analysis utilizes data from England for 2003–2005. The Annual Survey of Hours and Earnings is used to compute the SSWDs. This is a large national data set based on 1% sample of employees identified through the UK’s government (Pay as you Earn) tax system. This data is pooled over the 3 years to ensure the most precise estimates and reduce any year on year volatility. Vacancy rates were constructed as our measure of shortage of staff. This combines data on staffing levels with vacancy counts for positions which had been adver- tised for more than 3 months. This data was pro- vided by the Department of Health and is at hospital trust level. There are 209 hospital trusts (an administrative unit comprising of one or more hospital sites providing care to the local population) in each of the 3 years resulting in 627 observations.

The distribution of vacancy rates is presented in Table 1. On average, there is a vacancy rate of 1.2% for ANs and 2.5% for RNs. More than half of the hospital trusts do not have any vacant posts for ANs, while for RNs, this is only the case for 17% of hospital trusts.

IV. Empirical specification

SSWDs

The first stage in the empirical analysis is the construction of our measure of pay competiveness using the private–public pay gap. This is based on the estimation of SSWDs from four wage

Table 1. Vacancy rates distribution in hospitals, 2003–2005 (627 Obs.). Vacancy rates Mean SD NB Null P30 P50 P70 P90

Assistant nurses 0.012 0.028 53.27% 0 0 0.007 0.036 Registered nurses 0.023 0.032 17.38% 0.004 0.012 0.023 0.062

6548 J.-B. COMBES ET AL.

equations. Two represent the two public sector staff (AN, n = 2993 and RN, n = 10,761) and two represent their comparators in the private sector (AN comparator, n = 11,896 and RN com- parator, n = 29,041) as in Eq. (2):

lhik ¼ αþ X0βþ μk þ εi (2)

where lhik is the log of hourly earnings of individual i who works in area k. The matrix X contains the standardizing variables (age, age-square, gender, year dummies and for the private sector only indus- try and occupational dummies), εi are the indivi- dual-specific error terms and μk are the area-specific effects: the SSWDs. The SSWDs are estimated at Local Authority District Level. This is an adminis- trative unit in the UK with over 300 areas defined in England at the period of our study. They are esti- mated using effects coding, applied to Eq. (2) and finally centred to create our two public–private pay gap measures for ANs and RNs. The process of centring means that a region with a positive AN (RN) public–private pay gap has an AN (RN) SSWD relatively larger than the private sector com- pared to the national mean difference. Conversely, a negative public–private pay gap indicates an area where the AN (RN) SSWD is smaller relative to the private sector compared to the national mean differ- ence. The value of the gap indicates by how much this difference is in percentage terms. For example, a gap of 0.1 for RN indicates that the standardized RN wage in the area is more competitive by approxi- mately 10% compared to the national mean.

Vacancy model

The model to be tested is the following:

VACANCYj ¼ /j þ β1gap AN k þ β2gap

RN k

þ X 0 jγþ εj (3)

where VACANCYj is the vacancy rate of hospital j for ANs (RNs) and β1 and β2 are the parameters for gapANk and gapRNk which are our constructed mea- sures for the public–private pay gap of ANs and RNs for area k and our main parameters of interest.

X is a matrix of control variables and includes size, type of hospital trust, number of hospitals in each trust and the trust’s foundation status. The mean number of beds is around 700 beds per

hospital with an SD of more than 400 beds. The type of hospital trust was identified from Crilly et al. (2007), where trusts are distinguished as Specialist, Acute, Teaching, Mental Health or Other Trusts. Just over half of hospital trusts are classed as acute with around 5% classed as Other (children or specialist such as orthopaedic hospi- tals). During the period, an additional Foundation Trust (FT) status was rolled out. FTs were granted more freedom from central control including the ability to manage their own budgets. FT status was granted on a selection basis by the government based on financial and quality aspects of perfor- mance. Around 13% of hospital trusts had or were about to be awarded FT status in the time period studied.

The distribution of the vacancy rates would suggest estimation of Eq. (3) by a Poisson zero inflated model. However, after testing this spe- cification against OLS, we found similar mar- ginal effect for all variables. We therefore present the simpler linear model estimated by OLS. We estimate Eq. (3) with a modified variance-covariance matrix to take into account the repeated hospital trusts within the data set. This provides cluster-robust SEs (Arai 2011; Cameron and Miller 2010).

V. Results

Table 2 reports the summary results of our con- structed measure of the public–private pay gap for ANs and RNs.

It is this measure that is included within the vacancy model (3) to test whether the pay gap is linked to the both shortage of own staff and if it impacts on other staff. The results for both ANs and RNs are reported in Table 3.

There is no evidence that the type of hospital trust has any impact on the vacancy rates of staff. Neither is there evidence that the number of beds is associated with vacancy rates. Hospital trusts with foundation

Table 2. Public–private pay gap. N Mean SD Min Max P10 P50 P90

Assistant nurses gap

627 0 0.123 −0.33 0.4 −0.16 0.003 0.127

Registered nurses gap

627 0 0.111 −0.28 0.245 −0.16 0.007 0.129

APPLIED ECONOMICS 6549

status have a lower vacancy rate for RNs. However, our focus of attention is the public–private pay gap measures. The results indicate that the RN vacancy rate is associated with RNs pay competitiveness as measured by the RN public–private pay gap measure. There are lower vacancy rates in hospital trusts for which the RNs pay is more competitive. This is con- sistent with Elliott et al. (2007), (2010).

There is no effect of the AN public–private pay gap on the shortage of ANs. However, the vacancy rate of ANs is negatively associated with the RNs pay gap. A more competitive RNs pay increases the supply of ANs. Areas where pay for RNs is more competitive by 10% compared to the national mean decreases the RNs vacancy rates by 0.59% and decreases the ANs vacancy rates by 0.44%. Effects are therefore around 15% and 18% of the SD for AN vacancy rates and RN vacancy rates, respec- tively. There is no evidence that the pay competi- tiveness of ANs impacts on the vacancy rate of RNs.

VI. Discussion and conclusion

This article tests for the impact of local pay com- petitiveness on nurse staff shortages using a mea- sure of the public–private pay gap of ANs and RNs. Where hospitals cannot adjust the rate of pay to reflect local labour market conditions, they may face difficulties to attract and retain

staff. In particular, we allow for interactions between these two types of nurses and test for both the effect of the public–private pay gap for ANs on the shortage of RNs and vice versa.

We find no evidence that the pay competiveness of ANs influences their own vacancy rate. This might be considered surprising as we might expect that with low levels of specialized skills, ANs may be able to directly transfer to private sector occupa- tions offering more competitive pay. However, the absence of such an effect may reflect other rewards such as stability of employment and pension schemes that the public sector offers and this may be relatively more important for lower educated workers (Van De Walle, Steeijn, and Jilke 2015).

Our results confirm previous findings that the RNs pay competitiveness drives down vacancies of RNs. In addition, RNs pay competitiveness also drives down vacancies of ANs. ANs are more easily recruited in hospitals with more competitive RNs pay. This link may relate to skill-mix and task substitution between the two types of nurse. Task delegation from regis- tered to ANs is widely performed in hospitals within the UK (Cavendish 2013).Where there are difficulties in hiring RNs, more may be expected from the ANs on the ward and this may affect their recruitment and retention.While there is much written in terms of the impact of staff workload on issues such as nurse burnout and intentions to quit relating to RN beha- viour (Kutney-Lee et al. 2013; Aiken, Sermeus, and Vanden Heede et al. 2012), there is less written on ANs. Lacher et al. (2015) consider the impact of the working environment on the behaviour of healthcare assistants who provide support to RNs within the Swiss healthcare system. They found a negative asso- ciation between a self-reported working environment measure (that included staffing and resource ade- quacy as one of five elements) and measures of burn- out and intentions to quit.

A second link between the two nurse types may be through career-ladders. If the position of AN is regarded as a bridge to the more qualifiedRN, then the attractiveness of the RN position may influ- ence the shortage of the entry-stage AN position. However, there is no published evidence as to how many ANs do transition to fully RN in the UK and as such this link is more conjecture. There is some past evidence that suggests hospitals did support such career-ladders through secondments to

Table 3. Vacancy rate regressions for nursing staff. Assistant nurses Registered nurses

Estimate (SE) p-Value

Estimate (SE) p-Value

Intercept 0.01 (0.004)

0.011** 0.023 (0.005)

<0.001***

Gap AN −0.013 (0.014)

0.361 0.005 (0.018)

0.792

Gap RN −0.044 (0.02)

0.024** −0.059 (0.021)

0.006***

Number of beds 0.001 (<0.001)

0.862 −0.001 (<0.001)

0.866

No. of hospitals −0.001 (0.001)

0.817 −0.001 (0.001)

0.994

Foundation Trust −0.003 (0.003)

0.223 −0.007 (0.004)

0.068*

Mental 0.011 (0.008)

0.197 0.006 (0.008)

0.449

Teaching 0.001 (0.006)

0.872 0.002 (0.006)

0.768

Other 0.008 (0.014)

0.561 0.009 (0.016)

0.564

Specialist 0.001 (0.006)

0.832 0.005 (0.011)

0.667

Number of observations 627 627 Adjusted R-square 0.050 0.0418

***p < 0.01, **p < 0.05, *p < 0.1.

6550 J.-B. COMBES ET AL.

training (Grimshaw 2009). More recently, the role of AN as a potential entry point to full registered status has been recognized in Switzerland. Trede and Scherwi (2013) find evidence that the inten- tions of ANs to progress to RN is indeed influ- enced by future incomes. Interestingly within the UK, a new role of associate nurse has been recently developed which includes a ‘pathway’ to allow progression to RN status.

It should be noted that while the results of this article indicate recruitment of ANs may be easier in areas where RN pay is competitive, evidence suggests care must be taken in substituting one group for the other. Research that specifically looks at the skill-mix of RN and AN finds that substituting ANs for RNs may impact negatively on patient outcomes (see Aiken, Sloane, and Bruyneel et al. 2014; 2017; Griffiths et al. 2016).

Finding evidence that cross correlations of the competitiveness of pay exist between nursing staff groups implies that further research should inves- tigate if they exist elsewhere in the healthcare sector. One extension is to link the pay competi- tiveness of nursing staff with the vacancy rates of doctors and vice versa. If the public–private sector pay gaps captures hidden characteristics of the workplace, finding for instance an impact of doc- tors pay competitiveness on the RN shortage, we would expect it to be the consequence of increase workloads, and vice versa. This potential spillover effect suggests that measures to tackle one occupa- tional shortage may require consideration of the competitiveness of pay of other occupational groups.

A shortage of healthcare staff is recognized as a global problem against ever increasing demands on health services. As such it would seem beneficial to better understand the impact of the competitiveness of pay between the alter- native occupational groups that make up the healthcare workforce. This could provide some insight into both the economic drivers and potential solutions in terms of pay reform to support the supply of the healthcare workforce.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the Medical Research Council under grant number G0800113-2.

ORCID

Jean-Baptiste Combes http://orcid.org/0000-0001-6594- 8458

References

Arai, M. 2011. “Cluster-Robust Standard Errors Using R.” Stockholm University. http://people.su.se/~ma/clustering.pdf.

Cameron, A., and D. Miller. 2010. “Robust Inference with Clustered Data.” In Handbook of Empirical Economics and Finance, edited by A. Ullah and D. Giles, 20103666 vol, 1–28. Chapman and Hall/CRC. doi:10.1201/b10440-2.

Cavendish, C. 2013. Cavendish Review : An Independent Review into Healthcare Assistants and Support Workers in the NHS and Social Care Settings. London, UK: Department of Health. https://www.gov.uk/government/uploads/system/uploads/ attachment_data/file/236212/Cavendish_Review.pdf.

Crilly, T., J. Crilly, M. Conroy, R. Carr-Hill, and D. Parkin. 2007. Review of Specific Cost Approach to Staff Market Forces Factor. London: Department of Health.

Elliott, R. F., A. Ma, A. Scott, D. Bell, and E. Roberts. 2007. “Geographically Differentiated Pay in the Labour Market for Nurses.” Journal of Health Economics 26 (1): 190–212.

Elliott, R. F., A. Ma, M. Sutton, D. Skatun, N. Rice, S. Morris, and A. McConnachie. 2010. “The Role of the Staff MFF in Distributing NHS Funding: Taking Account of Differences in Local Labour Market Conditions.” Health Economics 19 (5): 532–548. doi:10.1002/hec.1489.

Elliott, R. F., M. Sutton, M. Ada, A.McConnachie, S. Morris, N. Rice, and S. Diane. 2006. Reviewing the Market Factor Forces Formula. London: Report to the Department of Health.

Griffiths, P., J. Ball, T. Murrells, S. Jones, and A. M. Rafferty. 2016. “Registered Nurse, Healthcare Support Worker, Medical Staffing Levels and Mortality in Englsih Hospital Trusts: A Cross-Sectional Study.” BMJ Open 6: 2.

Grimshaw, D. 2009. “Can More Inclusive Wage-Setting Institutions Improve Low-Wage Work? Pay Trends in the United Kingdom’s Public-Sector Hospitals.” International Labour Review 148: 4.

Kutney-Lee, A., E. S. Wu, D. M. Sloane, and L. H. Aiken. 2013. “Changes in Hospital Nurse Work Environments and Nurse Job Outcomes: An Analysis of Panel Data.” International Journal of Nursing Studies 50 (2): 195–201.

Lacher, S., S. Geest, K. Denhaerynck, I. Trede, and D. Ausserhofer. 2015. “The Quality of Nurses’ Work Environment and Workforce Outcomes from the Perspective of Swiss Allied Healthcare Assistants and Registered Nurses: A Cross-Sectional Survey.” Journal of Nursing Scholarship 47: 458–467.

APPLIED ECONOMICS 6551

Metcalf, D., K.Hansen, andA.Charlwood. 2001. “Unions and the Sword of Justice: Unions and Pay Systems, Pay Inequality, Pay Discrimination and Low Pay.” National Institute Economic Review 176 (1): 61–75. doi:10.1177/002795010117600106.

Office of National Statistics. 2000. Standard Occupational Classification Volume 2. England: Palgrave Macmillan.

Rosen, S. 1986. “The Theory of Equalizing Differences.” In Handbook of Labor Economics, vol 1 edited by O. C. Ashenfelter and R. Layard, 641–692. Amsterdam: Elsevier.

Smith, A. 1776. An Inquiry into the Nature and Causes of the Wealth of the Nations. London: W. Strahan and T. Cadell.

Trede, I., and J. Scherwi. 2014. “Work values and intention to become a registered nurse among healthcare assistants.” Nurse Education Today 34 (6): 948–953.

Trede, J., and J. Scherwi. 2014. “Work values and intention to become a registered nurse among healthcare assistants.” Nurse Education Today 34 (6): 948–953.

Traxler, F., and B. Brandl. 2011. “The Economic Impact of Collective Bargaining Coverage.” In The Role of Collective Bargaining in the Global Economy, edited by S. Hayter. Cheltenham, UK and Northamption, MA, USA: Edward Elgar.

Van De Walle, S., B. Steeijn, and S. Jilke. 2015. “Extrinsic Motivation, PSM and Labour Market Characteristics: A Multilevel Model of Public Sector Employment Preference in 26 Countries.” International Review of Administrative Sciences 81 (4): 833–855.

6552 J.-B. COMBES ET AL.

Copyright of Applied Economics is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.