Pensions Institute

Portability and Preservation of Pension Rights in the U.K.>

Published as Vol. III of the UK Office of Fair Trading Director General's Inquiry into Pensions, July 1997.

ByDavid Blake and J. Michael Orszag, Birkbeck College, Univ. of London

Research disseminated by the Pensions Institute may include views on policy but the Pensions Institute itself takes no institutional policy positions.

PDF of Report (121 pages)

Executive Summary

For at least two decades, the issue of the pension rights of early leavers has been a matter of public policy concern in industrialized countries whose pension systems rely heavily on funded defined benefit private sector pension schemes. In the U.K., the average worker changes jobs at least five times in his or her career and only a small proportion of workers will hold the same job at retirement as when they began pensionable service. Yet about 60% of full-time male workers and 53% of full-time female workers are in an occupational pension schemes, most of which are of the defined benefit type in which pensions rights are either not fully portable or not fully preserved. We show in this study that these people can lose a substantial fraction of their pension rights because they switch between jobs with different pension plans. This capital loss leads to labour market imperfections which help to reduce job mobility.

In this study, we examine the ways in which early leavers from pension schemes are penalized because:

Pension rights are based on the final salary in each job, with limited revaluation for inflation. This will lead to lower benefits if there is real earnings growth over the life cycle.
Defined benefit plans involve implicitly backloaded contributions so that workers who move to defined contribution schemes in mid-career will lose pension benefits because they will not receive the benefits of the backloaded employer contributions in their late career.
If employees leave a scheme during the vesting period (currently two years), they will lose their pension rights.
There are penalties and variations associated with the calculation of transfer values to other pension schemes (including defined contribution schemes), or welfare costs of remaining in a defined benefit scheme when individual circumstances have changed.

In Chapter 2, we review the economics of pensions and portability. Our main conclusions are:

Private pensions provide tax incentives to encourage savings.
In the absence of more direct and precise instruments for dealing with unrecovered training costs, monitoring employee performance and inducing retirement in a world where the firm does not have complete knowledge concerning the type or ability of the workers it employs, the economics literature argues that the pension scheme will be used (however bluntly) to deal with these matters.

In Chapter 3, we review the U.K. legislation on portability and explain how the rules affect the computation of transfer values and deferred pensions. Our conclusion is that the outcome of U.K. legislation over the past two decades has been a dramatic improvement in the position of early leavers. Portability losses are still large, however. In addition, the transfer values that workers receive are subject to a large degree of actuarial discretion that can dramatically affect their size. Recent legislation (in particular the Pensions Act of 1995) has actually had the perverse effect of reducing transfer values for early leavers.

Chapter 4 quantifies the portability losses faced by workers who change jobs. Our analysis focuses on two types of portability loss:

Cash equivalent losses arise because the early leaver's leaving salary is revalued to retirement age at a less favourable rate than used to determine the projected final salary. In computing transfer values, actuaries use the `current unit method with revaluation' to revalue salaries, while they use the `projected unit method' to project final salaries. These different methods lead to fewer `added years' being credited in a new scheme than are earned in the leaving scheme. We show that, other things being equal about the entering and leaving scheme, about the jobs and about actuarial projections, the only factors determining the cash equivalent loss are the ages at separation and the estimated real growth rate of wages. The cash equivalent loss is largest in absolute terms for workers who leave schemes in middle age, but is relatively the highest for the youngest early leavers.
Backloading losses due to the implicit backloading of contributions in a defined benefit scheme causes additional losses to those who switch to schemes which do not backload contributions. This type of loss is an especially significant consideration with the increasing importance of money purchase schemes.

From 6 April 1997, the 1995 Pensions Act imposes a Minimum Funding Requirement (MFR) for defined benefit pension schemes contracting out of SERPS. The individual's equivalent of the aggregate Minimum Funding Requirement (MFR) is the Minimum Cash Equivalent. For the Minimum Cash Equivalent component of transfers, actuaries are required to base their assumptions concerning real wage growth, inflation rates and yields on securities on the MFR norms as specified in Guidance Note 27 of the Faculty and Institute of Actuaries, although they have some discretion to deviate from these norms. For any deferred benefit in excess of the Minimum Cash Equivalent, actuaries have considerably more discretion.

We show that, even if the MFR norms are realized and the actuaries make all transfer value calculations based on these norms, early leavers experience at least the cash equivalent loss. In addition, they also experience the backloading loss if their new scheme is a defined contribution scheme with age-independent contributions. With the MFR norms realized, early leavers will be indifferent as to whether to leave a deferred pension or to take the cash equivalent of their deferred pension as a transfer value to their next scheme.

Leaving a deferred pension is a better option if the actuaries overestimate the degree of future real wage growth and a transfer value is a better option if actuaries underestimate the degree of future real wage growth. The MFR norms stipulate a constant growth rate in wages of 2 % per annum which is close to the historical average in the U.K.. However, it is a stylized fact of modern labour markets that most workers experience higher real wage growth early in their careers than nearer retirement. The implication of this is that the MFR norms will in general overestimate future wage growth for early leavers and as a consequence bias down the number of `added years' that a given cash equivalent buys in a new scheme. This means that it is preferable on average for early leavers to choose to leave deferred pensions rather than to take transfer values to their new scheme. The downside to the financial advantage of the deferred pension option is the administrative inconvenience of having to draw a number of deferred pensions in retirement.

In Chapter 5, we apply a number of realistic job separation histories to the lifetime earnings profiles of a range of `average' and `typical' U.K. workers. We show that cash equivalent losses can be quite substantial: between 10% and 20% of the full service pension for those choosing deferred pensions and up to 30% for those taking transfer values. Losses are significantly larger for those switching into schemes that do not benefit from the implicit backloading of contributions in defined benefit schemes.

In Chapter 6, we review the discretion available to actuaries in making their calculations of cash equivalents. We compute elasticities of actuarial discretion which measure the ratio of the percentage change in the computed pension benefit to the percentage change in the actuarial assumption. We evaluate these elasticities at the MFR norms and find that small changes in actuarial assumptions can have relatively large effects on the value of the pension in payment. We examine the following actuarial discretionary parameters in detail: discount rates, inflation uprating factors, annuity factors, and wage growth rates. With the exception of the annuity factor, the impact of changes in different assumptions depends on the time to retirement. We also identify a number of other areas (such as the valuation of discretionary benefits) over which the actuary also has discretion and on which it is not possible for us to comment in the absence of published documentation as to standard practice. We accept that individual circumstances can vary tremendously and recognize that either actuarial discretion or more complex rules (beyond the MFR assumptions) might be needed.

Chapter 7 examines pensions portability in other countries. We find that the Netherlands offers the highest degree of pensions portability and the U.S., at present, the least. The UK, along with Canada and Japan, comes somewhere in between.

In Chapter 8, we present a policy proposal which would not require major changes in legislation but would still improve dramatically the treatment of early leavers. It involves workers receiving some of their accrued contributions back when they leave a scheme in addition to a fraction of their accrued service credits calculated according to current methods. However, this proposal does not eliminate the portability losses of early leavers, it only reduces them. Within the context of defined benefit schemes, full portability requires either the complete transferability of service credits or the complete indexing of deferred pensions to real wage growth. In the absence of these changes to current practice, full portability in the context of private sector schemes can only be achieved using defined contribution schemes.

Remarks, comments and contributions are welcome at: dblake@econ.bbk.ac.uk.