Adjusting Pay-as-you-go Financed Pension Schemes to Increasing Life Expectancy

Schmähl, Winfried; Viebrok, Holger Article Adjusting Pay-as-you-go Financed Pension Schemes to Increasing Life Expectancy Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Schmähl, Winfried; Viebrok, Holger (2000) : Adjusting Pay-as-you-go Financed Pension Schemes to Increasing Life Expectancy, Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 120, Iss. 1, pp. 41-61, https://doi.org/10.3790/schm.120.1.41 This Version is available at: https://hdl.handle.net/10419/291949 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Schmollers Jahrbuch 120 (2000), 41-61 Duncker & Humblot, Berlin Adjusting Pay-as-you-go Financed Pension Schemes to Increasing Life Expectancy* By Winfried Schmahl and Holger Viebrok Abstract This paper deals with the impact of increasing life expectancy on pay-as-you-go (PAYG) financed pension schemes and measures that aim at solving the resulting budgetary problems. Relevant determinants for financial considerations are particularly the ratio of beneficiaries to contributors and the pension level. The focus is on measures that directly integrate indicators of life expectancy into pension calculation. The authors discuss the effects of a) a general reduction of the pension level by introducing a life expectancy indicator into the pension formula, b) a reduction of the initial average pension at a given retirement age or c) an increase of the retirement age itself because of increasing life expectancy. The authors conclude that postponing retirement is an adequate measure for coping with the financial effects of increasing life expectancy Zusammenfassung Der Aufsatz behandelt die Auswirkungen der steigenden Lebenserwartung auf umlagefinanzierte Alterssicherungssysteme wie zum Beispiel die gesetzliche Rentenversicherung in Deutschland und mögliche Maßnahmen, um den entstehenden Finanzproblemen zu begegnen. Für die finanziellen Auswirkungen sind insbesondere das Verhältnis von Rentnern zu Beitragszahlern und das Rentenniveau bedeutsam. Im Zentrum der Betrachtung stehen vor allem Maßnahmen, die die steigende Lebenserwartung direkt in die Rentenformel integrieren. Die Autoren diskutieren die Auswirkungen a) einer allgemeinen Senkung des Rentenniveaus, b) einer Veränderung der Rentenabschläge in Abhängigkeit vom Rentenbeginn sowie c) einer Anhebung des Rentenalters bei steigender Lebenserwartung. Die Autoren folgern, dass die Heraufsetzung der Altersgrenzen eine angemessene Antwort darstellt, um auf die finanziellen Konsequenzen steigender Lebenserwartung in umlagefinanzierten Alterssicherungssystemen zu reagieren. JEL-ClassifiCation: J14, J26, H55, J18 * This paper has benefited from comments by three anonymous referees. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 42 Winfried Schmahl and Holger Viebrok 1. Introduction Besides fertility and migration, mortality is one of the three determinants of the level of population as well as of its development and structure. In this paper, we are dealing solely with the impact of increasing life expectancy on pay-as-you-go (PAYG) financed pension schemes and discuss measures coping with the resulting budgetary problems. This special focus, however, should not be misinterpreted as a statement on the widely discussed question of PAYG versus capital funded pension schemes. Funding is already an important characteristic of occupational as well as of personal pension schemes beside PAYG schemes in many countries, like Germany For example, the German government decided to attach more importance to supplementary schemes, in particular to the third tier (private old age provision), whereas the statutory old age pension scheme shall remain PAYG financed. Such a shift usually implies distributive effects among cohorts, because more time is available for younger cohorts to accumulate appropriate savings. Though this subject is not within the scope of this paper, it belongs to its general conditions. Because all tiers have to fit into a comprehensive concept for the development of the old age security system, we place our focus on the impact of certain policy options in the PAYG financed part of old age security for different cohorts. We also refer to reform measures that have been decided on or implemented in several countries aiming at a reduction of public expenditure in old age security. Some of these measures try to take the increasing life expectancy explicitly into consideration and introduce indicators for life expectancy directly into the pension scheme. This will be discussed in detail. We mainly compare the following two alternatives and some of their effects: 1. Reducing the general pension level for all cohorts at a certain point in time, 2. Cohort-specific adjustments, especially a) cohort-specific pension levels or b) a method of linking the retirement age to the increase in life expectancy. We only need a simple model to remind us of some basic interrelationships. German demographic indicators are used to illustrate some of our arguments, but this should be considered as an example only. In order to clarify the relevance of the German figures, we place the demographic situation into that of the other OECD-countries at the beginning of section 2. Our conclusions, however, are more generally related to earnings-related PAYG Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 43 financed insurance schemes, in which benefits at least depend on the length of the working period. Such a scheme may be a 'defined benefit' scheme as well as a 'notional defined contribution' (NDC) scheme (Cichon 1999), but a direct link between contributions themselves and benefits is not essential for our arguments. 2. Demographic Changes 2.1 Changes in the Old Age Dependency Ratio Old age dependency ratios are often used as an indicator for an increasing 'burden' in aging populations. Figure 1 shows a remarkable increase in this ratio (with age groups defined at 65+/15-64) between 1960 and 1997 for many of the OECD Countries. It is well known that the reasons for this increase can be found in two processes: the rapidly declining fertility rate in particular at the end of the 1960s and the decreasing mortality especially of older persons, resulting in higher life expectancy.1 In addition, the rates are affected by the consequences of the two world wars. In order to get some more generalizable figures, we use an imaginary static population as a basis for further considerations in the following. It is constructed exclusively from age-specific mortality rates. Therefore we assume that mortality by age and sex and the number of birth remains constant over time or rather fertility is exactly as high as necessary to sustain the number of women and men within the population. These assumptions fulfilled, cross-sectional as well as longitudinal analyses lead to the same frequencies within the age-groups. In the next step, we derive several old age dependency ratios from that population. The result is shown in Figure 2, based on the life table 1995 / 97 for Germany. In Figure 2, the age on the x-axis is used as the borderline between the 'elderly' and the 'younger' people within the population, i.e. between the age groups in the numerator and denominator of the old age dependency ratio. Graph (1) shows the old age dependency ratio that is calculated by using age groups X+/ [15 up to (X-l)], where X indicates the age on the x-axis.2 Unlike graph (1), graph (2), starting at age 60, indicates dependency ratios with a fixed number of people in the denominator, taken from the total of 1 For a short overview, see Daykin/Lewis 1999. 2 To provide an additional explanation: Figure 2 is the result of moving the vertical line in Figure 4 below to the right and evaluating the quotient of the area below the curve on the right side divided through the one on the left side at every point on the x-axis. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 44 Winfried Schmähl and Holger Viebrok 11997 Dl960 30% Country Source: OECD 1999, p. 6-7 (database: OECD Labour Force Statistics). Sorted by ratio in 1960 * defined as 65+/15-64, Italy: 14-64, Sweden: 16-64; Belgium, Greece, Iceland: 1996 instaead of 1997; Germany 1960: West Germany only. Figure 1: Old Age Dependency Ratios* in OECD Countries Age Source: Own calculations based upon data of official federal statistics (Statistisches Bundesamt 1997a) Figure 2: Old Age Dependency Ratios within a Static Population, calculated at different Age-Limits, according to the German Life Table 1995 / 97 Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 45 age groups (15 up to 59), i.e. using age groups X+/(15 up to 59). If all people within these age groups participate in the old age pension system, graph (2) would indicate the systemic dependency ratio, if working life and hence the contribution payments stop after age 59. This could take place, for example, if an existing special pre-retirement scheme bridges the gap between the end of the working period and the beginning of the pension. The resulting difference between (1) and (2) at age 65 is 3.5 percentage points for women and 2.4 percentage points for men. This is relatively small compared to the expected development of the old age dependency ratio until 2030 (see below). We can see by the comparison of ratios in Figure 1 with ratios in Figure 2 that mortality (in Germany) leads to relatively high dependency ratios even in a static population. Under these assumptions the old age dependency ratio defined at age 65 (i.e. age classes 65+/15-64) would be 24.8% for men and 34.3% for women (see Figure 2, graphs (1)). Although these ratios are higher than the empirical ones today, they are substantially lower than predicted for the future.3 That future increase is caused by declining mortality as well as by low fertility rates since the end of the 1960s. It is not necessary to emphasize here that the development of the population structure is only one of the determinants that affect the budget of a pension scheme. For financial considerations, the ratio of the number of beneficiaries (pensioners) to contributors ('pensioner ratio', 'systemic dependency ratio') is decisive. Especially the labour force participation rates and their changes result in important differences in both variables. But also the development of average pensions and average wages are significant for the financial conditions of the pension scheme (for details see below). The question whether a higher number of pensioners results in higher pension expenditure depends, amongst other things, on how pensions are calculated and whether the pensions are adjusted to the development of (average) wages or not.4 3 Figure 1 above shows for Germany a combined ratio of 24% for men and women in 1997. According to own calculations (based on data of Enquete-Kommission 1998: 114), the old age dependency ratio at age 60 (60+/15-59) will increase from 38% (women and men) in 2000 up to 61% in 2030 and 63% in 2040, whereas the stable ratios at age 60 are 36.7 and 48.1, respectively (see also Figure 2). 4 For Germany, the association of the German statutory pension insurance agencies (VDR) calculates a so called 'standardized pensioner ratio' in order to eliminate the effects of such structural changes. This ratio is expected to rise from 42.5 in 2000 up to 60.6 in 2030. This is an increase of 43%, whereas the demographic old age dependency ratio is expected to increase much more in Germany if one uses the definition (60+/15-59) of the ratio (see above). According to the German Ministry of the Interior (Bundesministerium des Innern, without year) this ratio will increase from 40.9% in 2000 to 73.2% in 2030, an increase of about 80%. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 46 Winfried Schmahl and Holger Viebrok 2.2 Developments in Mortality and Life Expectancy Investigating the effects of increasing average life expectancy, it is important to know in which age segment a decrease of mortality takes place. For a long time, an increase of the average life expectancy (at birth) was achieved - amongst other things - by a decrease in infant and child mortality. In recent time, an increase in life expectancy is above all a result of a decrease of mortality rates in old age. For example in West Germany, the life expectancy at age 65 has increased relatively much more than the (average) life expectancy at birth (see Table 1). This increase in the life expectancy of the elderly has mostly been underestimated in the past. Table 1 Life Expectancy at Birth and at Age of 65 in West Germany 1960-1995 Life table At birth At age of 65 Life table Males Females Males Females 1959/60 66.69 71.92 12.38 14.37 1993/95 73.53 79.81 14.79 18.57 Change in percent 1960/95 10.3 11.0 19.5 29.2 Source: Enquete-Kommission (1998: 45, table 14). Especially during the past two decades, the increase in further life expectancy for the elderly was remarkably higher than in the previous period since the beginning of the twentieth century. This is shown in Figure 3 as an example for the life expectancy of persons aged 65.5 In West Germany around 1960, the life expectancy for women aged 65 was 79.6 years on average, the further life expectancy thus was about 14.5 years. 5 These data are based on cross-sectional mortality tables and refer to West Germany (1949/51, without the federal state 'Saarland', however). - Life expectancies in East Germany were and are below those in West Germany. Changes also arose in the situation after the unification: "A comparison of mortality in the East and West of Germany prior to and after the unification shows a divergence of life expectancy for men up to age 50, while the differences in life expectancy for women decreased"; (Sommer 1996: 18). A comparison between East and West Germany for the period 1950-1995 is given in Enquete-Kommission (1998: 46, Fig. 5). In general, an approach of the life expectancies in East and West Germany is expected due to the increasing similarity in living conditions. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 47 It increased by additional 4 years to 18.6 years until the mid-nineties. Life expectancy for men is clearly lower. Moreover, the difference in (average) life expectancy for men and women even increased during the recent decades. Around 1960 it was 2.2 years, while in the middle of the nineties the life expectancy of women aged 65 was then 3.8 years more than that of men. Life table Source: Statistisches Bundesamt (federal statistical office) 1997b: 76. Figure 3: Development of Life Expectancy at Age 65, West Germany Not only the increasing life expectancy in general but also the higher difference in the life expectancy for females and males affects the financial situation of pension schemes, because not only (ceteris paribus) the average duration of contributors' pensions increases - but also of survivors' pensions. Hypotheses about the future development of mortality and life expectancy are rather controversial. As already mentioned, deaths more and more shift into higher adult age (the curve of survivors increasingly approaches a rectangular form). "Populations are aging, but even given current age distributions, about one-third of male deaths and half of female deaths in developed countries occur after age 80" (Vaupel/Lundstrom 1996: 278). Therefore, in highly developed national economies a further increase of life expectancy is based on the decrease of mortality at higher age. Two developments have to be considered (see Figure 4): (1) Is there a decrease in mortality in the upper age range with the maximum life span reSchmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 48 Winfried Schmähl and Holger Viebrok maining largely unchanged or (2) is there a shift in the upper 'limit' for the maximum life span as well? In the latter case, the number of very aged persons (e.g. persons aged 80+) would also increase. With regard to the question as to what extent an extension of the life span may be assumed, different hypotheses are supported in the literature.6 Figure 4: Changes Resulting in Higher Life Expectancy Increasing life expectancy for the elderly is assumed in the population prognoses, for example, for Germany. The assumptions are mainly orientated toward the fact that in many other highly developed countries a significantly higher life expectancy has already been reached. On the other hand, one should remember that expectations concerning the future are often strongly shaped by the developments of the recent past. In a long-term perspective - e.g. for growth processes in biology - for many variables, there is often a development with a at first slow, then accelerated increase, which later decelerates again (S-shaped development). As can additionally be seen from Figure 4, the effects of postponing retirement age (shift of the vertical line) on the number of retirees and people at working age (areas below the curves) depend particularly on mortality at retirement age (i.e. the slope of the curve at retirement age). Effects of changing the retirement age are discussed below. 6 For discussion, see Vaupel / Lundström (1996) and the literature given there, as well as Manton, Stallard and Corder (1998), Enquete-Kommission (1998: 48 f.) and Birg and Börsch-Supan (1999: pp. 72 - 79). In the population projections of the United Nations the maximum life expectancy has increased by 10 years within 15 years; see Enquete-Kommission (1998: 49). Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 55 average net earnings (here symbolised by the first fraction). An additional factor based on the life expectancy at age 65 (X) was included, however with a time lag of 8 years and only with half of its weight. In the next years, this would have reduced the adjustment rate by 0.5 percentage points per year. Obviously, this reduces the pension level compared to an adjustment rate based on average net earnings development.14 The suspended adjustment formula (6) consisted of several arbitrary elements: • The monthly pensions of all cohorts are affected to the same extent, even if their own life expectancy did not change. • The division by two has no systematic reason, just like the time lag and the introduced floor (all are the result of a compromise on different political objectives concerning the pension level). As mentioned above, the demographic factor was substituted by an immediate ad-hoc cut of the pension level during 2000 and 2001. In these two years, pensions are adjusted to the inflation rate instead of net wage growth. The first rate was expected to be lower. 4.4 Cohort-Related Standard Pension Age If life expectancy increases and retirement age remains constant, the span for receiving pensions is extended in absolute and relative terms. Ceteris paribus, the number of pensioners, the pension expenditure and the necessary amount of revenue increases. In principle, the total amount of pension payments for a cohort between retirement age and the end of the cohort's life-span is (amongst other things) determined by the total number of remaining years of all cohort members until the end of the life-span and by the pension level. If the total number of remaining cohort years is increasing, this can be compensated by an adjustment of the initial average pension and the deductions at a given retirement age or by an increase of the retirement age itself. This is similar to the calculation of initial pension amount in the individual perspective described in Figure 5, but in order to calculate the financial effects of postponing retirement age exactly, we have to take into account that the changing number of 14 Nominal pension cuts because of the demographic factor were prohibited, thus the formula would only have become effective, if the average net earnings increase was at least as high as the cut. There was also a floor for the pension level introduced, i.e. the demographic factor would have been suspended as soon as this level had been reached. Because of that guarantee and the uncertainty of life span development it was hard to predict, how long this transition process towards the lower pension level would last (surely longer than one decade). Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 56 Winfried Schmahl and Holger Viebrok pensioners is not only due to the shift of retirement age but also to mortality at retirement age (see also Figure 4) and the conditions of eligibility for survivors' pensions. Hence a refined actuarial analysis is necessary to determine the appropriate deductions for a certain national pension scheme. That is beyond the scope of our paper. LE = Life expectancy A9e Sum of annuities= EUR 200.000 Figure 6: Compensation for Pension Level Reduction by Prolonging the Working Period How an increase of life expectancy leads to lower annuities based on the assumption of fixed total payments was shown in figure 5. Left aside was the fact that prolonging working life results in higher pension claims in an earnings-related pension scheme, due to the rules of calculating pensions. This can under certain conditions partly cover the reduction of the pension level. This is shown in Figure 6, which is based on Figure 5. Here a new graph (through (C")) exemplifies the annuities that are possible if we take into account the additional pension claims (and contributions) derived from extending the working period. As can be seen, in order to maintain the monthly pension level it is enough to postpone the retirement age from age (A) to age (C") instead of age (B), i.e. to a smaller extent. If life expectancy is increasing, a reallocation of time in the life cycle by extending working life and not by extending the time spent in retirement to the full amount of an increase in life expectancy seems to be plausible: "... there is a natural merit in extending retirement ages and discouraging earlier retirement" (Chand/Jaeger 1996: 31). Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 57 However, if retirement age is flexible, the effective retirement age depends not only on the standard retirement age but also on labour market, health and household conditions, income, and individual preferences.15 It is possible to increase standard retirement ages based on ad hoc decisions as it is done in many countries. Such an increase is mostly phased in over a long time. Besides ad-hoc increases in standard (and maybe also minimum) pension ages, it could also be based on a rule that takes life expectancy development into account. Such a rule could be: if life expectancy increases, retirement ages will be raised, too. This can take place not only for the standard retirement age - i.e. age for taking up the full pension without actuarial deductions - but also for the minimum retirement ages. A possible reason is that otherwise the effect of actuarial deductions from the full pension in case of early retirement reduces the absolute amount of pension benefits 'too much' according to distributive objectives. Such a rule was proposed for Germany as an alternative to a general reduction of the pension level (Schmahl 1997, 1998). To allow workers to adapt to new rules, such an increase in standard retirement age can be announced some years before it becomes effective, i.e. the increase in retirement age follows the development of life expectancy with a time lag of some years. If life expectancy rises, e.g. by one year, the increase in retirement ages will be less than one year. Calculating the distribution of one year of additional life expectancy (at age 65) between (a) an increase in working period and (b) an increase in additional time in retirement (to fix the new standard retirement age), two possibilities are mentioned for illustrative purpose: 1. Calculation in the PAYG-Scheme: assuming the aim is not to increase the contribution rate and not to reduce the average (gross) pension level, we start with a pension level I = ^ = 0.5, a contribution rate c = 0.2 and the pensioner ratio ^ = 0.4. Under these conditions, the increase of life expectancy by one year would increase NP, if retirement age remains constant. Based on these data the allocation of an additional 12 months (to leave the pensioner ratio constant at 0.4) lead to a 8.6 months higher standard retirement age as well as earliest age for initial eligibility for an old age pension. This would also result in 3.4 months of additional time in retirement. 2. Aaron and Reischauer (1998: 101) propose a rule for distributing additional years (after the phasing in of the standard retirement age in the 15 There exists a great number of publications dealing with the connection of mortality decrease with development of morbidity, e.g Manton et.al. (1997). Wise (1997) demonstrated by means of an option value analysis based on data of a great firm, that the behaviour also depends strongly on the benefits of occupational pensions. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 58 Winfried Schmahl and Holger Viebrok US from 65 to 67 in 20ll16: "Workers who spent the same proportion of their adult lives in retirement as those terming at 62 in 2011 would receive unchanged benefits. In other words, if retirement represented onefourth of adult life in 2011, a one-year increase in adult longevity would lead to the benefit cut associated with a nine-month increase in the age at which unreduced benefits are paid". To retire before the standard retirement age results in a lower pension because of (actuarial) deductions. If workers know the 'rule of the game' and can expect an increase in the retirement age, they are confronted with the decision to work longer and not to receive a reduction in pension or to retire earlier with lower benefits. 4.5 Resulting Distributive Effects among the Pensioners Several preconditions are necessary to implement an extension of working life in case of further increase of life expectancy: labor market conditions that allow workers to remain employed; retraining also of older workers; possibilities to 'save' for retirement (either as individual additional savings or via the firm as occupational pensions or on time-saving accounts) in order to have assets in case of earlier exit from the labour force. Savings may be used to bridge the time span between exit and taking up full pension or to compensate deductions in case of retirement before the standard retirement age. If the 'standard' retirement age is raised but the effective retirement age remains unchanged, eventually strong cohort effects occur among the pensioners for several years, because only certain cohorts, in particular those who retired after the reform and during periods with high unemployment rates, are affected. On the other hand, the fact that only future retirees are affected by postponing the retirement age fits well into a concept that includes a gradual shift from the first to the third tier. As mentioned at the beginning of this paper, such a shift eventually discriminates against older cohorts, if it is realised by a general reduction of the pension level, because older cohorts eventually will not have enough time to accumulate appropriate savings. (This should, however, be viewed in connection with the distributive effects of other measures within the cohort's life span.) In contrast to a general reduction of all pensions at the same time, the linking of retirement age to life expectancy does not affect older cohorts and younger people have the possibility to provide privately for their income in old age. This is the more im16 US Government had decided to phase this in by 2022. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 59 portant, the greater the pressure is to reduce the pension level for other reasons, e.g. because of low fertility rates. A second problem is, however, that the remaining high contribution rates eventually do not allow for additional private savings, especially in lower income groups. This may be solved by modifying the income tax rates. Postponing the retirement age cannot cope with all demographically induced financial problems of the pension scheme, like a declining fertility rate, and especially not with sudden shocks that reduce revenues. But it is well suited to deal with the financial consequences of the gradually increasing life expectancy, especially if the labor market conditions allow for compensating the deductions by extending the working period. 5. Conclusion Demographically induced financial problems in PAYG pension schemes are in the centre of pension reform debates in many countries. The reasons for a rising old age dependency ratio are mainly past declining fertility rates and currently rising life expectancy. The paper dealt with the latter, i.e. reduction of mortality. It was argued that the financial consequences of the increasing life span for a PAYG pension scheme should be compensated for by raising the standard retirement age (and eventually the earliest retirement age, too). If actuarial deductions (that make the system to a great extent independent from individual decisions on the effective retirement age) are implemented this would reduce the pension level if behaviour remains unchanged. But as was demonstrated in this paper, because of the additional contributions and claims resulting from extending the working period in an earnings-related pension scheme, the postponement of retirement age must not be as long as the increase of life expectancy. Additional years of one's life are shared between work and retirement. Several rules that aim at raising the retirement ages and ensure stable conditions for long-term decisions were presented in this paper. Whether they are suitable at all depends mainly on labour market and working conditions as well as human capital development, personnel policy and the institutional framework of individual old age provision. Furthermore, social policy has to consider the country-specific structure of the entire old agesecurity system, including the first tier and supplementary schemes. Linking the retirement age to the slowly increasing life expectancy fits into a concept that gradually puts more emphasis on capital funded supplementary schemes, because more time is available for members of younger cohorts to build up appropriate savings. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 60 Winfried Schmähl and Holger Viebrok Raising retirement ages, however, responds correctly to the increase in life expectancy, but not to declining fertility. Hence measures of this kind as discussed above are often a necessary, but not a sufficient condition for financial stability of pension schemes. In particular, the impact of unemployment and low fertility makes it harder to sustain the current pension level and may lead to further differentiated measures that could not be considered in this paper. However, postponement of the retirement age is also necessary for the future sustainability of PAYG pension schemes in aging populations. References Aaron, H. J./ Reis chauer, R. D. (1998): Countdown to Reform - The Great Social Security Debate, New York. Birg, H. /Borsch-Supan, A. (1999): Für eine neue Aufgabenteilung zwischen gesetzlicher und privater Altersversorgung, expert opinion for the 'Gesamtverband der deutschen Versicherungswirtschaf t', Bielefeld and Mannheim, mimeo. Bundesministerium des Innern (without year): Modellrechnungen zur Bevölkerungsentwicklung in der Bundesrepublik Deutschland bis zum Jahr 2040, Bonn. Chand, Ch. K. /Jaeger, A. (1996): Aging Populations and Public Pension Schemes, International Monetary Fund, Occasional Paper 147, Washington. Cichon, Michael (1999): Notional defined-contribution schemes: Old wine in new bottles? International social security review, 52, 4, 87 -106. Coronado, J. L./Fullerton, D./Glass, Th. (1999): Distributional Impacts of Proposed Changes to the Social Security Scheme, NBER Working Paper 6989, Cambridge. Daykin, C. /Lewis, D. (1999): A Crisis of Longer Life: Reforming Pension Systems, Institute of Actuaries and Faculty of Actuaries, Dublin. Enquete-Kommission "Demographischer Wandel" (1998): Zweiter Zwischenbericht, Bundestags-Drucksache 13 /11460, Bonn. James, E. (1998): New Models for Old-Age Security: Experiments, Evidence, and Unanswered Questions, World Bank Research Observer, 13, 2, 271-301. Manton, K. G. / Stallard, E./Corder, L. (1997): Changes in the Age Dependence of Mortality and Disability: Cohort and other Determinants, Demography, 34, 1, 135157. OECD (1999): OECD in Figures, Paris, OECD. Tables were obtained from http: //www.oecd.org/publications/figures/Index.htm (Dec. 1999). Rutkowski, M. (1999): The Quest for Modern Solutions: Pension Reforms in Transition Economies, mimeo. Scherman, K. G. (1998): Die Rentenreform in Schweden, Deutsche Rentenversicherung, 6-8/98, 413-431. Schmähl, W. (1987): Social Policies for Reducing Demographically Induced Costs in Social Security, European Journal of Population, 3, 439-457. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14 Pay-as-you-go Financed Pension Schemes 61 - (1989): Labour Force Participation and Social Pension Systems, in: P Johnson, C. Conrad, D. Thomson (eds.), Workers Versus Pensioners - Intergenerational Justice in an Aging World, Manchester/New York, 137-162. - (1990): Demographic Change and Social Security, Journal of Population Economics, 3, 157-177. - (1997): Alterssicherung - Quo vadis?, Jahrbücher für Nationalökonomie und Statistik, 216, 413-435. - (1998): Insights from Social Security Reform Abroad, in: M. J. Graetz, A. H. Munnel (eds.), Framing the Social Security Debate: Values, Politics, and Economics, Washington, 248-271 & 280-286. Schmähl, W. / Horstmann, S. (eds.) (2000): Transformation of Pension Systems in Central and Eastern Europe, Cheltenham. Sommer, B. (1994): Entwicklung der Bevölkerung bis 2040 - Ergebnis der achten koordinierten Bevölkerungsvorausberechnung, Wirtschaft und Statistik, 46, 7, 497503 & *438-*440. - (1996); Zur Entwicklung der Sterblichkeit aus der Sicht der amtlichen Statistik, in: R. H. Dinkel, Ch. Höhn, R. D. Scholz (eds.), Sterblichkeitsentwicklung unter besonderer Berücksichtigung des Kohortenansatzes, München, 11-25. Statistisches Bundesamt (1997a): Abgekürzte Sterbetafeln 1995/97, in: StBA, Fachserie 1, Reihe 1, 10.22. - (1997b): Statistisches Jahrbuch für die Bundesrepublik Deutschland, Stuttgart. Svensson, H. (1998): Swedish Pension System, in: Federation of Social Insurance Offices (eds.), The Future of Social Security, Stockholm, 192-203. Vaupel, J. WJ hundström, H. (1996): The Future of Mortality at Older Ages in Developed Countries, in: W. Lutz (eds.), The Future Population of the World - What Can We Assume Today?, International Institute for Applied Systems Analysis, Laxemberg. Viebrok, H. (1999): Wirtschaftswachstum, demographische Veränderungen und Weiterentwicklung der gesetzlichen Rentenversicherung, in: E. Grünheid, Ch. Höhn (eds.), Demographische Alterung und Wirtschaftswachstum, special issue of Schriftenreihe des Bundesinstituts für Bevölkerungsforschung, Opladen, 132-154. Wise, D. A. (1998): Retirement against the Demographic Trend: More Older People Living Longer, Working Less, and Saving Less, Demography, 34, 1, 83-95. Schmollers Jahrbuch 120 (2000) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.1.41 | Generated on 2023-04-04 12:27:14