پاداش کل و حقوق بازنشستگی در انگلستان در بخش های دولتی و خصوصی

Recent controversy has surrounded the relative value of public and private sector remuneration. We propose a comprehensive measure of Total Reward (TR) which includes not just pay, but pensions and other ‘benefits in kind’, evaluate it as the present value of the sum of all these payments over the lifetime and compare it for the highly educated in the UK public and private sectors. Our results suggest that TR is broadly equalised over the lifecycle for highly educated men while highly educated women have a clear TR advantage in the public sector by the end of their career. We suggest that the current controversy over public–private sector pension differentials and the perennial issues of public/private sector pay gaps requires a lifetime perspective and that the concept of TR is appropriate.

Recent controversy has surrounded the relative value of public and private sector remuneration and pensions in the UK. In the current recession and fiscal debt crises, there has been huge pressure to cut public sector remuneration. Many countries have already done this in nominal terms (e.g. Greece and the Republic of Ireland) and most countries will be doing this in real terms over the next five years. At the same time there has been growing concern about the ageing population and the burden of the pension obligations to public sector workers in the future. In the UK it is suggested by the Coalition government that public sector earnings and pensions are both too high relative to the private sector and therefore they both need to be cut in real terms. As any manipulation of public sector compensation (in terms of pay or pensions or other conditions of service) will have immediate consequences for fiscal budgets, workforce composition, delivery of services, inequality and relative remuneration it is necessary to carefully evaluate any proposed changes in any element of the total remuneration package. It is also important to be clear what this calculation tells us about public/private sector relative remuneration as this is a perennial comparison fraught with pitfalls. There is almost universal agreement that any debate about remuneration should include pay and pensions and all other forms of benefits in kind. There is no agreement on how this should be calculated. Although there has been a lot of work on selected aspects of the value of pensions across sectors (e.g., Disney et al., 2009) there has been relatively little on the evaluation of broader concepts of compensation. Indeed — although the notion of ‘Total Compensation’ or ‘Total Reward’ (TR) seems to have become widespread and fashionable in Human Resource Management circles there is no consensus of specifically what TR includes and leaves out. Often (see Greenhill, 1990 and Balsam, 2002) ‘Total Remuneration’ or the ‘compensation package’ (for executives) is said to include: salary, bonus, stock options, stock grants, pensions and other compensation. This literature tends to exclude: hours of work, holiday entitlements, job security (in terms of the probability of being made unemployed) and does not attempt to enumerate future benefits in present value terms or to adopt a life cycle perspective on this evaluation. These would all seem to be important considerations for an economic evaluation of TR. This paper provides a conceptual method for the measurement of TR and proceeds to estimate its structure for the private and public sectors in the UK. For the purposes of this paper we will define TR in a sector for an average career as the total financial benefits and ‘in kind’ compensation, evaluated in money terms over the life cycle. This will include conditions of work and all direct financial remuneration both now and deferred as pension payments in the future. Hence we take into account current earnings, pensions, hours of work, paid holidays, employer provided health insurance, the likelihood of unemployment and the lifetime pattern of pension contributions. We do this by pooling the largest available sources of data on public and private sector employees and examining how they differ, on average, across the life cycle. This means we use all of the following data in our analysis: the Annual Survey of Hours and Earnings (ASHE), the Labour Force Survey (LFS), the English Longitudinal Survey of Ageing (ELSA), and the British Household Panel Survey (BHPS). Each of these data sets provides different data on the various components of pensionable pay. We provide a Data Appendix to this paper which includes a list of all the available data which pertain to our evaluation of TR. The large literature on the earnings differentials in the public and private sectors (see Borland and Gregory, 1999) has addressed the issue of self-selection into careers and more recently considered the public-private differences in earnings dynamics (Postel-Vinay and Turon, 2007) over an employment lifetime. However this literature has completely ignored the value of pensions and the issues of TR over the complete life cycle (including retirement). This omission is very important in the light of the progressive changes in pension arrangements and the mooted reform of public sector pensions. The logical reasons for being interested in a dynamic model of sector choice are that one would wish to model: the propensity to self-select into either alternative on the basis of some unobservable characteristic (like ability or propensity to take risk), the potential for earnings variance to be different in the two sectors over the working life, and the possibility of modelling sector switching or mobility at different stages in the career. However, the price of solving the full dynamic programming model as in Postel-Vinay and Turon (2007) is that many simplifying assumptions have to be made — most notable of which is that individuals are assumed not to consider the non-pecuniary aspects of jobs like hours of work, the length of holidays and benefits in kind like medical insurance. But most important of all, this model cannot factor in the value of a pension into the employees' decision. Regrettably, this does not help us address the topical and fundamentally important question of whether public sector pensions are currently too generous. Indeed the only way to consider the whole issue is to develop a concept of TR. In the light of these considerations this paper proposes a new way forward. We take the group of highly educated who we consider can easily switch between the public and private sectors and so absolve ourselves of the self-selection issue of initial sector choice based on unobservable characteristics. (Our justification is that by restricting ourselves to graduates we can reasonably assume the individuals in our data to have a lower bound on unobserved ability and that risk in earnings profiles between the two sectors is not significantly different — as is borne out by public and private sector earnings variances over the employment lifecycle in Figs. A2 and A3). We also make the assumption that all graduates are risk neutral. Provided we make these limiting assumptions then we can get on and tackle the hugely important question of what TR looks like over the whole lifecycle and address the issue of whether pensions are too generous in the public sector. We can then also examine whether the pattern of switching sectors between the public and private sectors is consistent with what we observe about TR patterns. In other words, we ask whether there is a behavioural response to the inter-sectoral imbalance of lifetime TR in terms of the sector switching decisions that individuals make. Hence our perspective is the very real one facing governments all over the world — namely are public sector pensions too generous for the current labour force? This is not the same as asking how individuals facing occupation choices make their decisions and whether they switch sectors in a dynamically consistent way over their employment lifetime. The first contribution of this paper is to estimate the level of total compensation of the highly educated in the private and public sectors in the UK. The average earnings profile in the public sector depicted in Fig. 1 starts off at a higher entry level than in the private sector.1 Later in the life cycle stronger wage growth means that the private sector earnings profile rises above the public profile. While both profiles level off at later ages, the private sector profile even declines below the public profile. This shape of the private and public sector profiles has led researchers to impose a quadratic functional form on age-earnings profiles (cp. Disney et al., 2009). When performing the analysis on employer-reported earnings (ASHE data), we consistently find inverted u-shaped median age-earnings profiles (Fig. 1; the age-earnings profile using LFS data can be found as Fig. A12 in the Appendix).3 Basically, the question is whether initially low but steeper private sector earnings profiles produce the same TR as public sector profiles which (on average) start off higher but progress at a slower growth rate? To answer this question we define the concept of Accumulated Lifetime TR (ALTR). Besides earnings and pension accruals, we include four non-wage and non-pension components in the valuation of TR.4 So, the second contribution of this research is that we evaluate the contribution made to TR by: Hours of work, paid holidays, employer-provided health insurance and the probability of employment.5 After accounting for imperfections of the labour market (the risk of unemployment), the intuition is that TR in both sectors should be equal for very similar workers performing equal work. Therefore, in some sense, we perform this complex valuation exercise in order to assess a compensating differential over the lifecycle. The idea would be that each individual would attempt to choose the sector which maximised their lifetime TR (or utility).6 This may of course then involve switching from one sector to another at some point in the career. In such a framework it makes sense that the different alternative careers would end up having equal TR when calculated in present value terms over the whole lifecycle. If this was not the case then individuals would all wish to work in the same sector which would of course necessitate a realignment of at least one element of TR to bring the labour market back into equilibrium with appropriate amounts of people wishing to go into each sector.Our third contribution concerns the valuation of pensions. Unlike the earlier literature on pension valuation that computes the value of prospective one-year accruals in defined benefit (DB) and define contribution (DC) schemes (e.g., Disney et al., 2009 and Crawford et al., 2010), we account in greater detail for the complexities of the private and state components of the pension system in the public and private sectors. Specifically, we compute the level of total accruals at each age over the life cycle and compare a typical (‘average’) public sector employee (with more than 90% of DB coverage) with a private sector employee (with a mixture of DB, money purchase and state earnings-related pensions). This paper produces four empirical findings: firstly, we compute the value of non-monetary TR components in the UK at around 15 to 20% of total earnings — a non-negligible fraction. Secondly, the TR profiles of the two sectors cross for women who are better off in the public sector for most of their lifetime. For men, the private sector offers higher rewards almost until retirement, when lifetime TRs in the public and private sectors become equalised. This finding suggests some support for the lifecycle version of the ‘equalising differences’ story but also raises important questions about how early-career remuneration might affect graduates' sector choice. Thirdly, we find that the level of TR differs substantially across the public and private sector for most of the life cycle. The fact that total compensation is so different even after accounting for earnings, pensions, fringe benefits, work load as well as the risk of unemployment, has direct implications for the self-selection of employees across sectors. Finally our empirical work suggests that the behavioural pattern of revealed career switching is consistent with our evaluation of TR in the sense that men who start in the private sector switch into the public sector in greatest numbers exactly when the value of the pension kicks in — that it to say in their mid-forties. Likewise women who start in the private sector are always likely to switch into the public sector at an increasing rate throughout their career and there is very little movement in the other direction, i.e. from the public sector to the private sector. The remainder of this paper is structured as follows: Section 2 outlines the methodology of evaluating the TR packages in both sectors. An overview of all data sets that will be employed and all TR components is provided in Section 3. Section 4 presents the results on TR over the life cycle. Section 5 then evaluates whether the observed sector switching behaviour of workers in the data is consistent with our evaluation of TR. Section 6 concludes.

This paper evaluates the TR of highly educated employees in the public and private sectors in the UK across the life cycle. It provides for the first time a comprehensive measure of various monetary and non-monetary work related benefits in addition to earnings and pension accruals. The analysis suggests that there is not equality of TR profiles between the two sectors at every point in time. Yet, the ALTR for men is equalised between public and private sectors over the life cycle suggesting that the private sector earnings advantage at younger ages is counterbalanced by the more generous benefits associated with public sector pension schemes. This result implies that male university graduates who choose employment in either of the two sectors based on their potential early career reward prospects might get a biased signal with respect to lifetime reward. Women seem to be better off in the public sector at almost any point of the life cycle profile. If workers in both sectors were exposed to similar levels of workplace disamenities (e.g., stress or mortality risk) our results would imply that after taking into account pensions the public sector confers a high positive TR advantage for women but a very closely comparable one for men in the two sectors. This equalisation of total lifetime remuneration which, for men, balances the early career advantage of being in the private sector by the long run advantage of being in the public sector later in the career suggests that there may be a form of the compensating differentials theory for men in which the rewards of public service are currently balanced by those in the private sector. The paper also stresses the importance of benefits in kind and the role of workload in the valuation in TR. While a substantial literature deals with pensions as part of work-related remuneration, fringe-benefits, working hours and unemployment risk have not been studied comprehensively. Our results suggest that these employment aspects are important, and again more valuable in the public than in the private sector. Such a conclusion means that any discussion of public/private sector pay differentials or public/private sector pension differences should not be considered in isolation. What ideally should be calculated is TR in the two sectors as measured over the lifecycle. The behavioural consequences of the relative value of TR in the public and private sectors were evaluated in terms of the switching behaviour between sectors over the lifetime. We found that the pattern of graduate sector switching was consistent with the TR evaluation — namely that we saw men who initially choose the private sector shift into the public sector as their career advances and women were doing this progressively throughout their careers. An important implication of our results relates to the possible consequences of government reforms of public sector pensions for labour supply. If, as our results suggest, there is currently a broad comparability between public and private sector TR for highly educated men — then a substantial policy change which made public sector pensions less generous, would have a knock-on effect on public sector graduate recruitment and retention. A clear prediction of our results is that any move towards a Career Average pension scheme which was not compensated for in terms of higher earnings (or other benefits) would cause less people to enter the public sector, or leave it prematurely (or not switch to it in late career) hence inducing labour supply knock-on effects. Finally, it is important to stress that our conclusions regarding the evaluation of TR for all categories of non-graduates are very different to those set out here for graduates. This is essentially because non-graduate lifecycle earnings in the public and private sectors do not cross — specifically public sector non-graduate employees enjoy higher earnings throughout their working lives than their private sector counterparts. Hence the conclusions for pension reform are clearly different for this group. We examine the implications of our findings for a wider pension reform in further research.