- Shopping Bag ( 0 items )
Social Security Programs and Retirement around the World represents the second stage of an ongoing research project studying the relationship between social security and labor. In the first volume, Jonathan Gruber and David A. Wise revealed enormous disincentives to continued work at older ages in developed countries. Provisions of many social security programs typically encourage retirement by reducing pay for work, inducing older employees to leave the labor force early and magnifying the financial burden caused by an aging population. At a certain age there is simply no financial benefit to continuing to work.
In this volume, the authors turn to a country-by-country analysis of retirement behavior based on micro-data. The result of research compiled by teams in twelve countries, the volume shows an almost uniform correlation between levels of social security incentives and retirement behavior in each country. The estimates also show that the effect is strikingly uniform in countries with very different cultural histories, labor market institutions, and other social characteristics.
The Belgian social security systems face an uncertain future. One major reason is the financial burden imposed by the aging of the population. For the systems to survive this demographic process, higher contribution levels, lower benefits, or both will have to be introduced, given the pay-as-you-go (PAYG) nature of these systems. Indeed, a straight increase in the public debt to finance this demographic transition is not an option as it would mean pushing the already high ratio of public debt to gross domestic product (GDP) to even further astronomical heights. Most recently however, the successive Belgian governments have successfully brought down the debt-GDP ratio to close to 115 percent from a level of 130 percent by means of strict budgetary policy. The government has even achieved a small budget surplus in 2001 and has reached the same goal in 2002 even in the presence of an economic slowdown. This fiscal rigor will no doubt increase the margin of maneuverability of the federal government in its attempts to cope with the demographic aging process.
Another factor of uncertainty pertains to the consequences of increased labor mobility on the way the social security systems are organized. First of all, increased mobility between jobs in the public sector, in the private sector, and in self-employment may induce large changes in the way the three corresponding social security systems work. Recent reform proposals by the Belgian federal government to improve the way the public sector works also has to be seen in this light. Mobility between sectors will most likely increase once people's behavior is determined by similar factors measuring achievement and productivity in both the public and the private sectors. In those circumstances, a review and harmonization of the corresponding public retirement-income systems seems warranted. Second, the question of international job mobility is becoming more and more important, particularly for a small open economy in the heart of Europe like Belgium. Jousten and Pestieau (2002) study the implications of an expected increase in labor mobility from a European perspective, and the authors pay particular attention to the degree of redistribution inherent to the different systems. They argue that both levels of intra- and intergenerational redistribution will be widely affected, even if we replace the assumption of perfect labor mobility between member countries with a more restrictive and plausible one of mobility limited to individuals belonging to particular income groups.
However, even leaving these two challenges aside, the Belgian social security system needs reform. The widespread use of a variety of early retirement programs makes Belgium the country with the lowest average retirement age in the Western world, which is approximately fifty-seven years old for men. This chapter studies the incentives pushing people toward retiring early. We explicitly model the incentive structure built into the various public retirement and early retirement systems. First, we compute indicators of benefit entitlement, such as social security wealth. Then, we define several different incentive measures based on the notion of social security wealth. In a third step, we perform an empirical estimation of micro-econometric probit and option value models. From our exceptionally rich and broad database, we are able to compute a rather accurate measure of all individuals' pension wealth, as well as of the implicit tax rates the elderly workers face in case of delayed retirement.
The structure of the paper is as follows: Section 1.2 describes the essential features of the various public retirement and early retirement systems. In section 1.3, we explain the different, mostly administrative components of our large data set. The following section touches on the problem of the earnings process used in the simulations and estimations. Section 1.5 explains the process and logic underlying the construction of the different incentive measures used, while section 1.6 contains regression results obtained using these latter incentive measures. Section 1.7 delivers two policy simulations using the previously estimated coefficients. The first simulation consists of an increase by three years of the eligibility age in the various retirement systems, and the second consists of a policy in which early retirement would be possible at age sixty at the earliest, while normal retirement age would be sixty-five. Section 1.8 is devoted to the conclusions.
1.2 Social Security Schemes
The Belgian retirement income system rests upon three very unequal pillars. First of all, there are the public social security programs that represent the largest part of pension income for a wide majority in the population. The second pillar consists of company pension schemes, which only play a minor role as a source of income for the average Belgian worker. Essentially, they are currently confined to the higher-income individuals in the private sector and to the self-employed, a finding that is at least in part due to their tax treatment. A third type of retirement income comes from individual retirement savings. These take multiple forms: there are tax-favored individual-pension savings accounts with a maximum annual contribution of 580 [euro] per person (approximately U.S.$615), or there are more traditional savings vehicles, such as the tax-favored savings accounts, investments in trust funds, life insurance, and so forth. The dominance of the first pillar can also be represented in numbers: Whereas the first pillar represents pension entitlements of more than 250 percent of GDP, assets in private-pension funds only amount to 10 percent of GDP.
The first pillar, public retirement programs, essentially consists of four components. There are three large sectoral social security programs; one for the public sector, one for the private-sector wage earners, and one for the self-employed. Some special categories of workers, such as coalmine workers and military personnel, have special retirement systems that we will not analyze in the present paper. A fourth large category of public retirement income consists of the guaranteed-minimum-pension system that operates on a means-tested basis.
Aside from these pure retirement programs, the Belgian government has introduced early retirement provisions that either operate under the name of early retirement scheme, or alternatively as a form of old age unemployment. Table 1.1 gives a brief outline of the importance of the different categories of social security programs for the year 1995.
1.2.1 Wage Earners' Scheme
The wage earners' scheme is the largest program, according to the number of people affiliated with the program. The program allows for retirement between the ages of sixty and sixty-five, with the choice of the retirement age not inducing any actuarial adjustment. The system worked according to a different logic till the early 1990s. Until 1992, the wage earners' scheme had an actuarial adjustment factor of 5 percent per year of early retirement before the age of sixty-five.
However, in the case of most workers, the choice of the retirement age is not completely neutral with respect to the benefit amount, as most men still accrue extra pension rights by working additional years between the ages of sixty and sixty-five. This is so because a full earnings history consists of forty-five years of work for men, a condition that many people do not satisfy at the age of sixty. For those having more than forty-five working years, a dropout-year provision replaces low-income years with higher ones. The situation so far has been slightly different for women. Indeed, until very recently, women only needed forty years of earnings to have a complete earnings history. In reaction to successive court rulings on the illegality of this sex discrimination, the Belgian government introduced a reform a few years ago that aims at progressively increasing the complete career condition to forty-five working years for women over the time period 1997-2009. However, for most women included in our data set, a full career still consists of forty years of work.
Benefits are computed based on earnings during periods of affiliation. The benefit formula can be represented as follows:
Benefit = n/N x average wage x k,
where n represents the number of years of affiliation with the wage earner's scheme, N the number of years required for a full career (in our case either forty or forty-five), and k is a replacement rate, which takes on the value of 0.60 or 0.75 depending on whether the social security recipient claims benefits as a single person or as a household. The variable average wage corresponds to indexed average wages over the period of affiliation, with indexation on the price index combined with additional discretionary adjustments for the evolution of growth. A peculiar feature of the Belgian wage earners' scheme is that periods of the life spent on replacement income (e.g., unemployment benefits, disability benefits, workers compensation, and the like) fully count as years worked in the computation of the average wage, and hence of the social security benefit. For any such periods, fictive wages are inserted into the average wage computation. In line with the general philosophy of the Belgian social insurance system that any such spell on a replacement-income system is purely involuntary, imputed wages are set equal in real terms to those that the workers earned before entering these replacement-income programs.
An additional category of linked benefits is payable to surviving spouses or, more generally, to surviving dependents of deceased wage earners. All the different types of benefits provided for under the wage earners' social security system are covered against erosion by the means of inflation through an automatic consumer price index (CPI) adjustment.
The system works both with floors and ceilings, which are either indexed to the evolution of prices or to average wages. The minimum household pension represents a floor for workers that have contributed during their entire working life to the systems. It is approximately equal to 56 percent of average net wages. At the opposite extreme, a ceiling operates on pensionable but not on taxable earnings. The earnings entering the above pension formula are strictly limited to a maximum of 120 percent of average gross wages. Wage earners' pensions are also subject to an earnings test. Currently, the earnings limit is approximately 7,450 [euro], or $7,900, per year. For earnings above this limit, pension entitlement is suspended.
The wage earners' system is essentially based on the PAYG principle and financed through payroll taxes that are levied both on the side of the employers and of the employees with a combined contribution rate of 16.36 percent. The system also receives a subsidy from the Belgian federal government that is approximately equal to 11 percent of overall benefits.
Next to the official wage earners' scheme, several forms of early retirement programs have developed over the last few years, some officially carrying that name and others that do not (e.g., unemployment, preretirement, and so forth). Those schemes can be broadly subdivided into two groups: collective mandatory retirement and individual early retirement. During the 1980s and the 1990s, an arsenal of mandatory early retirement schemes was put in place. All of these arrangements were and are based on collective agreements, which are negotiated with the active involvement of employees and employers, sometimes at the sectoral level or sometimes at the level of an individual company or production site. For some companies in a difficult economic position, mandatory retirement ages as low as fifty were introduced. The federal government did not necessarily object to such arrangements, as it considered early retirement as a good tool in the fight against youth unemployment. Indeed, some of these early retirement schemes required the employers to rehire young workers. Lately however, these early retirement schemes have undergone some scrutiny. Not surprisingly, the beneficial labor market effects have been rather modest, if not completely absent. Recent discussions and decisions at the government level clearly move in the direction of lifting the effective early retirement age and hence also the sector-specific mandatory retirement ages.
This has to be seen in contrast to the massive costs these programs induce for the federal budget, as well as for the society as a whole. First of all, the effect of these waves of early retirement on the federal budget operates both through missing contributions during the period spanning from early retirement to the normal retirement age as well as through additional costs. Thus, the federal government pays a large fraction of early retirement compensation. On the other hand, an individual's pension rights in the wage earners' pension schemes are essentially unaffected by the decision to retire early or not. This is due to the previously discussed feature of the Belgian social security system that days spent on replacement income count as working periods in the computation of average pensionable earnings and of periods of activity. Hence, retiring early does not induce any loss of income during retirement.
Individual early retirement differentiates itself from its collective counterpart by the fact that it is based on an individual's decision to retire from work. The most prevalent way is to pass through the unemployment system in which the unemployed aged fifty or older are considered "aged unemployed," and thus neither subject to controls on availability to work, nor to benefit cuts due to long-term unemployment. Therefore, people unwilling to continue to work can ask their employer to lay them off. The latter has no incentive not to lay the worker off, unless the employer considers the employee to be a crucial wheel in the working of the company. Laying the worker off allows the employer to replace an expensive old worker by a cheaper young one. Furthermore, the employer's behavior does not add any costs to his unemployment contributions, as the system is not experience rated.
Next to the unemployment path, some people also attempt to proceed to retirement through the disability-insurance scheme. However, in the Belgian context, we think that disability is not a very prominent means of departure, at least not for private-sector employees. Incentives to claim disability benefits are rather limited: Medical screening is relatively severe, and benefits are not significantly more interesting than early retirement provisions.
1.2.2 Public-Sector Employees
The social security scheme for public servants is the oldest one of the three, dating back to as early as 1844. Public pensions are paid out of the general federal budget. Officially, the public sector pensions are considered as deferred income rather than old age insurance. The only official insurance aspects are the 7.5 percent payroll taxes that the public-sector employees have to pay to finance survivor benefits. Benefits are essentially individualized, that is, there are no additional spousal benefits available for no- or low-income spouses.
Civil servants' pensions are compulsory as of age sixty-five for both men and women. However, as for the private sector, there is a multitude of ways of retiring earlier than this normal age of sixty-five. It is possible to opt for an incomplete career and retire at sixty. For some particular categories of workers, the normal retirement age is lower than sixty-five, and early retirement provisions are sometimes extremely generous. This is particularly the case for military personnel and for teachers, who have always enjoyed a much more favorable treatment. For example, secondary-school teachers in the French-speaking community can either retire at age fifty-five if they have sufficient years of service, or alternatively take a less demanding route in terms of career requirements and retire at the age of about fifty-eight.
Excerpted from Social Security Programs and Retirement around the World Copyright © 2004 by National Bureau of Economic Research. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Introduction and Summary
Micro-Modeling of Retirment in Belgium
Income Security Programs and Retirement in Canada
The Impact of Incentives on Retirement in Demark
Estimating Models of Retirement Behavior on French Data
Micro-Modeling of Retirement Decisions in Germany
Micro-Modeling of Retirement Behavior in Italy
Social Security and Retirement in Japan: An Evaluation Using Micro-Data
Incentives and Exit Routes to Retirment in the Netherlands
Micro-Modeling of Retirement Behavior in Spain
Income Security Programs and Retirement in Sweden
Pension Incentives and the Pattern of Retirement in the United Kingdom
The Effect of Social Security on Retirement in the United States