Geographically differentiated pay in the labour market for nurses

Abstract

This novel application of spatial wage theory to health service labour markets analyses the competitiveness of nurse's pay and how this differs between local labour markets in Britain. A switching regression model is estimated to derive standardised spatial wage differentials (SSWDs) for nurses and their comparators. An SSWD gap is constructed and its relationship to vacancies estimated. A reduction in the gap in a local area is shown to result in an increase in the long-term vacancy rate for National Health Service (NHS) nurses. The competitiveness of nursing pay is shown to have a strong effect on the ability of the NHS to attract and retain nurses.

Introduction

Workforce issues are now a central focus of policy for the health services of many countries. A number of developed countries, for example, the UK, USA and Australia, face challenges in securing adequate numbers of motivated staff to deliver efficient health care while many developing countries face a challenge retaining their trained staff. The role of pay in influencing the recruitment, retention, and motivation of nurses is receiving increasing attention. Previous studies have examined the effect of the level of nurses’ pay on their labour supply, with labour supply defined in terms of hours of work and labour market participation (Antonazzo et al., 2003, Shields, 2004, Chiha and Link, 2003, Askildsen et al., 2003, Skåtun et al., 2005). The supply of labour is also determined by the relative pay offered in different jobs. The relative pay of nurses can differ substantially between local labour markets but this has not yet been examined in the literature. This study therefore draws on established theory from labour economics to construct and estimate a model which explores how variations in the competitiveness of pay between local labour markets affect recruitment and retention. It explores this for nurses and is the first study to do this for any occupational group.

In the UK health service the pay of the majority of health care professionals employed by the NHS is set centrally and exhibits little spatial variation. The competitiveness of pay therefore differs between professional groups and within professions between different regions and local labour markets. In his Budget statement in April 2003, the Chancellor of the Exchequer proposed introducing “measures to ensure that public service pay systems are more responsive to regional labour market conditions.” (HM Treasury, 2003). Nurses sell their labour in local labour markets and thus differences in the competitiveness of pay between local markets are of critical importance for this group of NHS professionals. The overall aim of the paper is to detail how the competitiveness of NHS nurses’ pay differs across local labour markets in Britain and to explore its consequences.

The competitiveness of pay is revealed by calculating the difference between standardised spatial wage differentials (SSWDs) estimated for nurses and their comparators. SSWDs reveal the ‘underlying’ structure of pay between different areas. To date SSWDs have only been estimated for the public and private sectors in the UK as a whole (see Blanchflower and Oswald, 1994, as reported in Elliott et al., 1996, and Wilson et al., 1996). When SSWDs are estimated for such broad groups of employees it is important to control for differences in the characteristics of the workforce in different areas: the human capital composition and the industrial and occupational mix. SSWDs are estimated using standard regression models and reveal substantial variation between different areas.

The theory of compensating wage differentials provides a theoretical framework to explain why the ‘underlying’ structure of pay, as measured by SSWDs, differs between geographical areas (Smith, 1776, Rosen, 1986). Competition in labour markets ensures that the net advantages of different jobs will tend to equality. Thus, we expect to find higher pay in some areas of the country where the cost of living is higher while higher pay is also necessary to compensate for a less pleasant working environment. Empirical research has provided support for this theory of spatial pay differences (Reilly, 1992, Shah and Walker, 1983, Blackaby and Murphy, 1995).

For the net advantages of jobs in different areas to be equalised, labour must be mobile, labour markets must be integrated and pay structures flexible. Where these conditions are not met, departures from equilibrium will occur. Disequilibria in labour markets may result from inflexible wage structures. Pay structures may reflect the preferences of those participating in the institutions that set pay. Where trade unions have an important role in pay setting, pay is likely to deviate from the rates that would otherwise be paid in the market. Trade unions are likely to be concerned about equity and fair pay, and often seek to negotiate a national rate for the job (see Metcalf et al., 2001). Where this happens they will narrow the distribution of pay and the resulting wage structure will be flatter than would otherwise occur. If the power of trade unions differs between areas (see Blackaby and Murphy, 1991) this will also affect geographical patterns of pay.

The incidence of trade unionism is much greater in the public sector than in the private sector. It therefore seems likely that the geographical pattern of pay in the public sector exhibits a much flatter distribution than in the private sector. Even though average pay may be the same in the two sectors, the private sector pay structure seems likely to exhibit greater dispersion than the public sector. This is illustrated in Fig. 1. The consequence is that the public sector will overpay in the low cost and high amenity area and underpay in the high cost and low amenity area.

This pattern would be expected for qualified nurses working in the NHS; in Britain the nurses union covers the whole of Britain and they are paid on national salary scales that exhibit very little spatial variation. In contrast the alternative occupations in the private sector in which nurses could work are likely to exhibit much greater spatial variation in pay. Any difference between the spatial distribution of nurses’ pay and that of their competitors will result in variations in the competitiveness of nurses’ pay between local markets. These variations in competitiveness can be captured by calculating the differences between the estimated SSWDs for these groups in each local labour market. Differences in competitiveness are likely to result in differences in the ability of employers in the health service in different areas to attract and retain nurses.

In the UK National Health Service (NHS) qualified nurses are paid on national, UK-wide, pay scales (see the Review Body for Nursing Staff, Midwives and Professions Allied to Medicine, 2004). Under these national scales the same terms and conditions apply across the UK and, except for London and the surrounding area, there are only minor differences in pay between different areas. Additional allowances are paid to those who work in London and contiguous areas to compensate for the higher cost of living and to offer some compensation for the disamenities associated with work in London (see the London Weighting Advisory Panel, 2002). However, beyond this employers in the NHS have little scope for aligning the pay of qualified nurses to conditions in local labour market conditions. Pay scales for these nurses are short and offer very little scope for either appointing new hires at different points on the scale or moving qualified nurses up to higher grades. The centralised pay setting arrangements for qualified nurses do not allow pay to be adjusted to address staff shortages in local markets.

NHS employers have some scope to adjust non-pecuniary benefits, such as the quality of the working environment, training and relocation expenses.¹ To meet their recruitment needs employers have participated in national initiatives, such as ‘return to practice’ schemes, which offer additional funding for re-training and recruitment of staff who have been out of the workforce for some time (Elliott et al., 2003). In addition some employers have recruited substantial numbers of nurses from overseas. Yet the persistence of nursing vacancies in some geographical areas and in some specialities suggests that more needs to be done to address these shortages.² The ‘Treasury view’ that public sector pay structures should be made more responsive to regional labour market conditions is one approach. In this paper we establish for the first time, using robust econometric methods, that the competitiveness of nurses’ pay differs between local labour markets. We further show that these differences in competitiveness are systematically related to staff shortages.

Our principal hypothesis is that the NHS nurse vacancy rate in a geographical area is determined by the difference between the relative pay of NHS nurses and that of the occupations in the same local labour market which are alternatives to NHS nursing. These occupations will be similar in terms of the skill they require and they may have features in common with nursing. In choosing between them nurses will evaluate the relative pay that they offer.

Our modelling strategy is therefore as follows. First we estimate a wage model which controls for the endogenous choice of nursing as an occupation. We establish the competitiveness of NHS rates of pay for nurses by estimating a switching model, comparing the pay of nurses and their comparators across different local labour markets in Great Britain using data from the Quarterly Labour Force Survey. From this we construct a measure of the competitiveness of nurses’ pay by health authority and then estimate how the differences in competitiveness affect the three-month vacancy rate for nurses in health authorities across Scotland, England and Wales. As the sample size becomes rather small when disaggregated by the geographical boundaries of health authorities, the relative competitiveness of nurses’ pay in some areas is likely to be inaccurately measured. In the vacancy estimation, we therefore control for attenuation that might be caused by the explanatory variable being inaccurately measured using two methods. The following section outlines the nature of the econometric model that is estimated. This is followed by a description of the data and variables used in the analysis. The results are then presented and discussed.