University of Amsterdam
Faculty of Economics and Business
The Shades of Grey
A study to create a more robust Dutch public pension
system
Master Thesis
Lisanne Bos (5873258)Abstract –
The discussion concerning pensions is a very delicate topic in the Netherlands. In this paper I focus on the first (public PAYG) pillar of the Dutch pension system. This first pillar suffers from the aging problem for three reasons. First, due to better living conditions, people are healthier and as a result they will live longer. Second, due to the baby boom generation, there has been an enormous increase in the number of people over the age of 65. A third reason is that the fertility rate isdecreasing. This paper discusses four different proposals in order to relieve the burden on the public finances: abolishing the pension, raising the pension age, creating a means-dependent AOW-pension system and making the AOW-AOW-pension depending on the number of working years. Increasing the pension age is the only proposal which can be implemented at this moment. Besides this, the government could also consider the proposal in which the AOW-pension is employment-related, and to find solutions in order to reduce the problems with moral hazard and insurance problems.
1 Introduction
3
2. The Dutch pension system
5
2.1. The three pension pillars
5
2.1.1 First pillar: AOW 5
2.1.2 Second pillar: Pension construction via the employer 6 2.1.3 Third pillar: Individual additional pension schemes 6
2.2 Summary
7
3. Aging problem
7
3.1 Population in the Netherlands
7
3.2 Dutch aging problem and Dutch public pension system
10
3.3. Empirical model
10
3.3.1 Dataset 12
3.3.2 The model 14
3.4 Summary
15
4. Creating a more robust Dutch public pension system
16
4.1 Changes made by Dutch government
16
4.2 Other options to make the Dutch public pension system more robust.
18
4.2.1 What if we abolish the AOW? 20
4.2.2 What if we raise the pension age even higher? 23 4.2.3 What if we make AOW income related? 27 4.2.4 What if we make AOW dependent on the number of working years? 30
4.3 Summary
32
5 Conclusion
34
6 References
36
1 Introduction
Changes in the Dutch public pension system is a trending topic. There is a television show, ‘Effe geen
cent te makken’ (penniless for a while), in which two ‘BN-ers’ (Dutch celebrities) try to live off of an
AOW-payment and it is a subject that is frequently written about in newspapers. An article by Yvonne Hofs (2013) published in The Volkskrant, ‘Hoezo oud en zielig?’ (Why old and pathetic?) lead to a agitated discussion. The Volkskrant never received so many reactions on an article ever before. The discussion concerning pensions is apparently a very delicate topic.
Looking at several papers written about the aging problem (Garssen, 2011; Knaap, Bovenberg, Bettendorf & Broer, 2003; etc.), there are three main sources found which create the aging problem. First, due to better living conditions, people are healthier and as a result they will live longer. Second, due to the baby boom generation, there has been an enormous increase in the number of people above the age of 65. A third cause is that the fertility rate is decreasing, as shown in graph 1. The fertility rate is the average number of children born to a woman in here lifetime. There are fewer children born per household, which means that the working cohort is decreasing.
Graph 1: Fertility rate in the Netherlands since 1935 (Source: CBS Statline, 13-01-2014)
As shown in graph 1, starting in 1935 there was an evident increasing trend in childbirth. Starting from the 70’s, there is an evident downward trend in the birth of children. This had to do with the introduction of birth control which was introduced by the feminist group called ‘De Dolle Mina’s’ who fought for women’s rights.
The Dutch pension system is known to be a strong and well regulated system. It consists of three pillars. The first pillar is the Dutch public pension which is called ‘Algemene Ouderdoms Wet’ (Social
40 50 60 70 80 90 100 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2011 2012 ◦⁄◦◦
Fertility rate
Security, General Law for Elderly or shortened, AOW). Since the AOW is a Pay-as-you-Go (PAYG) system, the working cohort needs to pay in order for the elderly to receive their pensions. The second pillar consists of a funded pension system by the employer. The third pillar is an individual tax-facilitated additional pension system. In this paper I focus on the first pillar of the Dutch pension system. The AOW guarantees a minimum income after retirement and is paid by the government (van der Lecq, 2010). This means that when the number of retirees rises, the total amount of pensions paid by the government rises as well, which imposes a considerable burden on long-term public finances (van Ewijk, Bettendorf, van der Horst, Draper, de Mooij, ter Rele, 2011). Due to this problem, questions about this system arise. Is this current system a sustainable system? What happens if we continue without any changes in the public pensions? Which changes will decrease the burden on public finances? Therefore, my research question is:
How can the Dutch public pension system be changed in order to relieve the burden on the public finances due to the aging-problem?
In order to answer to this question, I present a literature review and empirical research. I first describe the three pillars of the Dutch pension system: the AOW, the employer-funded pension system and the individual tax-facilitated additional pension schemes. Second, I look at the effect of the aging-problem in the current situation. I create an empirical model to examine the effect of the aging-problem on the total expenditures of the AOW made by the government. Using this model, I will show that the public pension needs to be changed due to the aging-problem. With the use of this model, I can clearly see the effect on the total government expenditures on the AOW when one of the variables changes. I explain this model in the third section. Thereafter, I will look at the current changes made by the government in order to relieve the burden on public finances and then I will look at several different proposals to improve the Dutch pension system even further. There is an interesting discussion (van Ewijk, et al., 2011; van der Lecq, 2010; Wolfsen, 2010; etc.) about the current situation and the reactions to the aging problem. The proposed solutions in these papers vary from abolishing the AOW, to creating a new system in which the AOW depends on the income during the working period and more (Wolfson, 2010; van der Lecq, 2010; van Ewijk, et al., 2011, Bos & Teulings, 2010; etc.). So, in my conclusion I will combine existing ideas and new insights in order to create a proposal to improve the Dutch public pension system.
2. Dutch Pension System
The Dutch pension system is known as a strong and well regulated system. In the first section of this paragraph, I explain the three pillars in the Dutch pension system in more detail, followed by a summary.
2.1 The three pension pillars
The Dutch pension system exists of three pillars: the first pillar is the public pension (AOW), the second pillar is a supplementary employer-funded pension and the last pillar consists of individual tax-facilitated savings account (Rijksoverheid, 2013).
2.1.1 The first pillar: AOW
The AOW is the first pillar of the Dutch pension system. Introduced in 1957, the AOW is a system that provides a basic income for elderly over the age of 65. Each person who lives or works in the Netherlands is entitled to AOW (Rijksoverheid, 2013). The AOW is a fixed amount and this amount is the same for each person regardless of their income during their working period. The amount of the AOW is set by the government on a yearly basis. There are certain criteria that need to be fulfilled to be entitled to the AOW. First, each person should have lived between 15 and 65 years in the Netherlands, if not then there is a 2% cut of the AOW for each year this person did not live in the Netherlands. Also, the amount of the AOW depends on the current situation of each person. If one is married or not-married but living together, and both have reached the age of 65, each of them will receive a pension which is 50% of the minimum income level. If the partner has not yet reached the age of 65, than the retired is entitled to an income-depended allowance. As shown in table 1, there is also a premium for people who live together but only one receives an AOW-pension.
Living Situation Amount (gross) of AOW per month*
Single €1,086.52
Living Together
- Both receive AOW-pension €750.93 - Only one receives AOW-pension
(without premium)
€750.93
- Only one receives AOW-pension (with full premium)
€1,476.70
Single parent €1,368.94
The level of this premium depends on the amount of years people are insured for AOW, the income of the partner and the total income of both (Sociale Verzekeringsbank, 2013). Singles are entitled to an AOW-payment which is 70% of the minimum income level and single parents are entitled to 90% of the minimum income level (Rijksoverheid, 2013). The AOW is financed according to a PAYG-system. This means that the working cohort is paying for the elderly through wage- and income taxes. Before 1998, the AOW was fully financed by means of a PAYG-system. The premiums were paid by the working cohort at a rate of about 17% to 18% percent of the first two brackets of income taxation (Kremers, 2002). After 1998, the premium tax was at its maximum level (18.25 percent of the first two income tax brackets), still the AOW costs were rising above this level. Therefore, not only the working cohort is paying for the AOW-pensions, because the contribution rate was maximized, the government was forced to contribute in order to compensate the financing of the AOW-pension scheme (Sol-Bronk & Vleeming, 2009; Kremers, 2002). Currently, the AOW-pensions are financed with two-thirds via the AOW-contribution only paid for by the working cohort and about one-third is paid via general taxes (Sol-Bronk & Vleeming, 2009).
2.1.2 The second pillar: Pension construction via the employer
The second pillar of the Dutch pension system is a pension construction via the employer. According to the Dutch government, about 90% of the employers offer their employees an additional pension system (Rijksoverheid, 2013). The second pillar of the Dutch pension system is fully prefunded through pension funds, taking into account an important degree of the prospective aging of the population (Kremers, 2002). Through this system, retired employees receive a supplementary allowance on top of their AOW-pension. The additional pension contributions are paid by the employer and the employees; usually two-thirds by the employer and one-third by the employee. In the Netherlands, firms are obliged to place the money for the additional pensions outside their own undertaking, this can be a pension fund or an insurer (Rijksoverheid, 2013). Dutch pension funds are among the largest in the world, taken together the first and the second pillars the collective old age income provision typically totaling a level of about 70 percent of gross earnings before retirement (Kremers, 2002).
2.1.3 The third pillar: Individual additional pension schemes
Individual insurances are the third pillar in the Dutch pension system. Each person is free to decide whether to use this pillar. Examples of this third pillar are: annuities, single premiums and life insurances. With these arrangements, one can save for an additional pension. Typically, contributions to these pension schemes and the savings account itself are exempt from income taxation and capital taxation. This third pillar can be used to fill a pension gap or to retire early (Rijksoverheid,
2013). Again, the Netherlands stands out with a relatively high level of accumulated savings (Kremers, 2002).
2.2 Summary
To summarize, the Dutch pension system consist of three pillars. The first one is called AOW-pension, this is for two-thirds financed by a PAYG funded pension, the working cohort is paying for the elderly, and one-third is financed by the government through general taxes. The second pillar is a pension construction via the employer, about 90% of the employers offer their employees an additional pension scheme. The first and the second pillar combined generate an old age income provision typically totaling a level of about 70% of gross earnings before retirement. The third pillar is an individual additional pension scheme via, for example, life insurances.
Because my thesis focuses on different options to relieve the burden on the public finances on pensions, I only focus on the first pillar.
3. The Aging Problem
The population structure in all Western countries is changing. The combined effect of a lower mortality rate, a lower birth rate and the baby-boom have a considerable effect on the structure of Western populations (Knaap, et al, 2003). As many governments promise pension and medical benefits to their elderly citizens, the burden of retirees’ benefits is becoming painfully obvious. Uncertainty about the future makes planning for retirement benefits even more difficult (Bohn, 2005). There are three main sources of the aging problem (Bohn, 2005; Garssen, 2011; etc.). First, due to better living conditions people are healthier and as a result they live longer. Second, due to the baby boom, there is a big rise in the number of people above the age of 65. Third, due to lower fertility rates, the working cohort is decreasing compared to the growth of retirees. In this section, I first describe the changes in the population in the Netherlands. Thereafter I discuss the aging problem in combination with the Dutch public pension system. In section 3.3, I will create an empirical model which shows the effects of the number of retirees on the total expenditures of AOW-pensions.
3.1 Population in the Netherlands
In the past half decennia, the Dutch population has grown every year (Garssen, 2011). The birth of children is by far the strongest contribution to the growth in the Dutch population. According to
Garssen (2011), between 1960 and 2010 there were over ten million children born and approximately six million people passed away. The natural growth of the population was therefore roughly four million, this growth has slowed down the aging of the Dutch population. This effect is even stronger in comparison with other European countries. Garssen (2011) claims that the Netherlands has a fast growing population due to the combination of relatively high birth rates and low mortality rates. Because the mortality risks of elderly are decreasing, the growth in the number of people above the age of 65 is determined by the birth numbers 65 years ago (Garssen, 2011).
Graph 2: Size and growth of the retired population (Source: CBS statline, 13-01-2014)
This means that starting in 2011 the growth in the amount of people above the age of 65 will increase, because from that year on, the first baby boomers turn 65, as can been seen in graph 1. As shown in graph 2, the raising trend of people above 65 already started in 2002, this can be explained due to the fact that before World War II, in 1937, there was already an increase in the number of births. Furthermore, as you can see in graph 2, during the last 50 years the aging in the Netherlands has been very gradual, except for the last years when the group of 65-plus grew with approximately 30 thousand people. 0 0,5 1 1,5 2 2,5 3 0 5 10 15 20 x mln
%
Size and Growth of the people above 65 years
Growth of the population (left) Size of the population (right)
0 5 10 15 20 0 2 4 6 8 10 x mln
%
Size and Growth of the population
Graph 3: Size and growth of the total population (Source: CBS Statline, 13-01-2014)
Another important fact is, that when you compare the growth of the total population in graph 3 with graph 2, you see that the population of people above 65 years grew much more gradual than the growth in the total population. This difference has to do with migration. Migration only has a minor role at higher ages (Garssen, 2011). Even though the developments move gradually, the composition of the population is changing. Notice, that if you compare graph 2 with graph 3, the growth of people above 65 years is much higher compared to the total growth of the population. This means that, as you can see in the graph below (graph 4), the old-age dependency ratio is increasing.
Graph 4: Number of elderly per amount of youth. (Source: CBS Statline, 08-12-2013).
Another problem is that not only the number of elderly is increasing but that they also have a lower mortality.
Graph 5: Average age of death in the Netherlands. (Source: CBS Statline, 08-12-2013).
Due to better living conditions and improving medical solutions people are healthier and as a result they live longer, as can been seen in the graph above (graph 5). This increasing group of elderly ensures that the demand for care and services for the elderly is raising.
0 5 10 15 20 25 30 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2011 2012 2013 %
Old-Age dependency ratio
50 60 70 80 90 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2011 2012 A ge
Average age of death
In the next section I will describe the consequences of the changing population on the Dutch public pension system.
3.2 Dutch Aging Problem and the Dutch Public Pension System
According to Knaap, et al, (2003), the situation in the Netherlands seems to be better compared to other countries who are also facing the aging-problem. This is because in the Netherlands, about 40% of retirement income comes from the second pillar. Therefore the pension contributions are less vulnerable to demographic changes. After all, the elderly have put money aside for their retirement and depend less on the contributions paid by the working generation (Knaap, et al, 2003). Though, due to the fact that the mortality rate is declining and the number of people in the working cohort is decreasing, the old-age dependency rate increases. Therefore, the costs of the AOW increase (Knaap, et al, 2003; van de Lecq, 2010; etc.).
As seen in graph 4 in the previous section, the old-age dependency ratio is increasing. This means that the affordability of the Dutch public pension system is under pressure. It is of course a positive idea that we live longer, but for the government it is a financial problem (Bohn, 2005). The AOW contribution is maximized since 1998, whereby the AOW-pensions are paid increasingly by general resources due to the aging-problem (Sol-Bronk & Vleeming, 2009). Because the AOW is a PAYG-system, the increasing proportion of the elderly becomes a serious problem. As the working cohort is decreasing, the government needs to react in order to be able to offer AOW to their elderly. In section four, I will discuss several ways in which the Dutch government can react to this urgent problem. In the next section, I illustrate the effects of the aging-problem on the expenditures of the AOW.
3.3 Empirical Model
In this part of my thesis, I demonstrate that the expenditures of the public pensions depend on the factors that play a role in the aging-problem. I use a simple empirical model that shows the effect of a change in the number of retirees on the total expenditures on AOW. The models used by the CPB are based on different assumptions, I have chosen to create my own model with additional assumptions in order to show the effect of aging on the total expenditures of the government on AOW-pensions. Models created by the CPB assume, for example, fixed amounts of AOW-pensions, while this is not the reality as can been seen in the following graphs.
Graph 6: Yearly amount of AOW-pension (Source: Towers Watson and CBS Statline, 13-01-2014). The amount of AOW received per year is not a fixed amount but, as you can see in graph 6, it increases and changes almost every year.
Graph 7: Amount of AOW-pension corrected with consumer price index (Source: Towers Watson and CBS Statline 13-01-2014).
Even if these numbers are corrected with the consumer price index (graph 7), you can see that the amount of AOW is not a fixed amount. Also, the ratio between the minimum wage and the AOW-pension has changed over time. As can been seen in graph 8, the ratio has changed from 70% and 50% to almost 100% respectively 70%. The rise was a result of the increase in the minimum wage and the new legislation in the area of tax and social security contributions in accordance with the proposals of the Commission Oort (Digibron, 1989). The proposals of the Commission Oort were an integration of national insurance contributions and income tax, a reduction in tax allowance, and a reduction in the number of income tax (van de Stadt, Huigen & Zeelenberg, 1989).
0 5 10 15 20 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Eu ro' s (x 1 0 0 0 )
Amount of AOW
AOW (married) AOW (unmarried)
0 5 10 15 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Eu ro' s (x 1 0 0 0 )
Amount of AOW
corrected with CPI
Graph 8: Amount of AOW as a percentage of the minimum wage (Source: Towers Watson and CBS Statline, 13-01-2014). Also, the growth of AOW doesn’t follow the same growth structure as the average income as you can see in graph 9. In the period between 2002 and 2008, the growth of AOW was much more gradual compared to the growth in average income. This has to do with the economic growth at that time, it was the period after the internet bubble, therefore average income increased more rapidly compared to social security payments.
Graph 9: Growth of AOW and Average Income (Source: CBS Statline, 27-01-2014).
I first explain the dataset, thereafter I will show the created model, followed by a summary.
3.3.1 Dataset
For this empirical model, I will use publically available data from the Statistics Netherlands and from Towers Watson. The AOW is intended for everyone above the age of 65, independent of someone’s income during his or her working period. The level of the AOW is determined every year by the government. Therefore the total expenditures of AOW depend on the number of people above the age of 65. The level of AOW is linked to the level of prices, therefore I use the consumer price index.
0 20 40 60 80 100 %
Amount of AOW as a percentage of the Minimum Wage
AOW (married) /Minimum Wage AOW (unmarried) / Minimum Wage
95 100 105 110 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 %
Growth of AOW and Average Income
Also, the level of AOW is linked to the level of minimum income, therefore I also include that variable in this regression. These variables are used for a regression to control whether they affect the total amount of AOW. The basic model for the regression looks as follows:
Where:
represents total expenditures on AOW, in euro’s per year at time t;
represents the total number of retired people in the Netherlands, per year at time t; represents the consumer price index (1990 = 100), per year at time t;
represents the monthly minimum wage, per year at time t.
Graph 10: a) Yearly total expenditures on AOW made by the government, b) Total number of 65+, c) Consumer Price Index (1990=100), d) Monthly minimum wage (Source: CBS Statline, 08-12-2013).
The data of the total expenditures on AOW ( ) is yearly data from 1960 to 2012. The data of the number of retired people ( ) is also yearly data from 1960 to 2012. The consumer prices index
( ) is yearly data from 1960 to 2012, the year 1990 = 100. The data on minimum wages ( ) is collected via Towers Watson. It is yearly data from 1975 to 2012. It shows the minimum income per month. I am using first differences because, as can been seen in the appendix, the variables are not stationary, using first differences makes these variables stationary.
Average Median Maximum Minimum Std. Dev.
(x mln) 593 593 1490 -671 438
32645 28804 121422 19853 15623
23 22 81 -28 22
3 2 7 -1 2
Table 2: Descriptive statistics of the data.
As can be seen in the table 2 and in the graphs 10, all of the data have increasing trends. According to the literature, the number of retirees should have a positive effect on the total expenditures of AOW. More retirees mean more pensions that need to be paid by the government. Also, an increase in consumer prices should imply a higher total expenditure on AOW, because if prices of goods increase, people need a higher income to be able to consume the same amount of goods. At last, an increase in the minimum wage should also imply an increase in the amount of total expenditures on AOW.
3.3.2 The model
As can been seen in the appendix, in section seven, the variables are not stationary, therefore I have taken the first difference of all the variables in order to make them stationary.
Dependent variable: Sample: 1976 2011
Number of observations: 36 (after adjustments)
Coefficient t-Statistics C -0.009319 -0.591373 1.683606 2.146949* 0.002298 0.768989 0.000884 4.420515* = 0.626246; Adjusted = 0.591207 *significant p-value (p-value < 0.05)
Table 3: Results of the OLS-calculation with the dependent variable and the independent variables , and .
Table 3 shows the results of the OLS-calculation with the dependent variable and the independent variables , and . I first discuss the results of the regression. The first difference of the total number of retirees in the Netherlands ( has a positive coefficient, which is as expected. It means that an increase of one percent in the number of retirees increases the total expenditures on AOW pensions by 1.68 percent. This variable is significant, as can be seen in table 3, assuming a significance level of five percent. This coefficient is probably more than one percent because the amount retirees receive depends on their state of living (married, single, etc.). The first difference in the consumer price index ( ) has a positive coefficient, as predicted in the literature before. If the consumer price index increases by one percent, the total expenditures of AOW will increase by 0.2298 percent. However, this variable is not significant, as can be seen in table 3, assuming a significance level of five percent. The first difference in the level of minimum wages ( ) also has a positive coefficient, as predicted in the literature. It means that, if the minimum wage increases by one percent, the total expenditures of AOW will increase by 0.0884 percent. This variable is significant, as can been seen in table 3 assuming a significance level of five percent. Apparently, the increase in minimum wage has a significant effect on the increase of the total expenditures on AOW and the consumer price index does not. A possible explanation is that the minimum wage is already taken into account the consumer price index, therefore an increase in minimum wage already implies an increase in the consumer price index.
3.4 Summary
The population structure in most Western countries is changing due to the combined effects of a lower mortality rate, a lower birth rate and the baby-boom. These three factors are the three main sources of the aging problem. In the half past decennia the Dutch population has grown every year, as can been seen in graph 3. The birth rate of children is by far the strongest contributor to the growth in the Dutch population. This growth has slowed down the aging of the Dutch population, in fact, this effect was even stronger compared to other European countries. Even though the developments move gradually, the composition of the population is changing. The old-age dependency ratio is increasing, which means that the number of elderly is raising faster than the number of young people. Since the AOW-pension scheme is a PAYG-system, the increasing old-age dependency ratio is becoming a serious problem for the financeability of the AOW. As the working cohort is decreasing, the government needs to respond in order to be able to offer AOW-pensions to their elderly without exploding contribution rates. The model created in paragraph 3.3 shows that a rise of 1% in the number of retirees will lead to an 1.68% increase in the total expenditures on AOW by the government. As can been seen in graphs 2 and 3, the retired cohort is increasing more rapidly compared to the growth of the population. This means that the total expenditures will increase
rapidly. Note that the AOW contribution has already been maximized in 1998 therefore the AOW is increasingly paid by general taxation. This means that, if nothing happens to the current regulations, the expenditures will increase enormously due to this aging problem. According to the model it will increase by 1.68% with every one percent increase in the number of retirees. As can be seen in graph 2, the growth of retirees was about 18%, this means an increase of almost 30% of the total expenditures on AOW-pensions.
In the next paragraph I discuss some options in order to create a more robust public pension system in the Netherlands.
4. Creating a more robust Dutch public pension system
The aging problem has social-economic consequences, according to Goudswaard (2002), the financial implications will have the biggest consequences. As shown in the model created in section 3.3, the expenditures on AOW-pensions will increase rapidly as the retired cohort keeps increasing. Other studies have also shown that the expenditures on pensions and healthcare will increase enormously due to this aging problem if nothing happens to the current regulations (Goudswaard, 2002). According to Goudswaard (2002), the increase in costs, due to the AOW, will be about 4% of GDP. The calculations in the study of Ewijk, et al (2011), show that if the government maintains its current GDP-share of revenues and expenditures in the future, the implicit debt included in future primary deficits equals about 300% of GDP. This is more than four times the current debt ratio of 65% of GDP. This amounts to an annual sustainability gap of about 4.5% of GDP, which indicates that sooner or later an adjustment to the current policies will be needed (Ewijk, et al,2011; Goudswaard, 2011). Because this paper focuses on the public pension system, I will find solutions to sustain the current public pensions. In the next sections, I show what the current changes made by the Dutch government are. After that, I look at four different options to change the public pension system in order to relieve the burden on the government expenditures on pensions. First, I will look what will happen if the government abolishes the AOW, followed by a discussion on raising the pension age. Then, a discussion about a means-related pension and finally I will look at the effects of making the AOW independent on someone’s income during his or her working period.
4.1 Changes made by the Dutch government
Currently, the government has decided to increase the pension age from 65 years up to 67 years (Kamerstukken II, 2009/10). This is done step-by-step because it gives people the chance to adjust to
the new situation. Not only for employees it is good to know beforehand when they are able to retire, but also for employers. Employers have taken into account the departures of older workers at the age of 65, so their policies should be changed due to an increase in the pension age. As shown in table 5, the retirement age is already increased in the first month of last year (Rijksoverheid, 2013). In the following years, the retirement age will be raised in small steps until 2019, then the retirement age is 66 years.
Born: Will receive AOW
in the year:
Age of receiving AOW:
Before 1 January 1948 2012 65
After 31 December 1947 and before 1 December 1948 2013 65 + 1 month
After 30 November 1948 and before 1 November 1949 2014 65 + 2 months
After 31 October 1949 and before 1 October 1950 2015 65 + 3 months
After 30 September 1950 and before 1 August 1951 2016 65 + 5 months
After 31 July 1951 and before 1 June 1952 2017 65 + 7 months
After 31 May 1952 and before 1 April 1953 2018 65 + 9 months
After 31 March 1953 and before 1 January 1954 2019 66
After 31 December 1953 and before 1 October 1954 2020 66 + 3 months
After 30 September 1954 and before 1 July 1955 2021 66 + 6 months
After 30 September 1954 and before 1 July 1955 2022 66 + 9 months
After 30 June 1955 and before 1 April 1956 2023 67
After 31 December 1956 2024 Unknown *
*The age to receive AOW-pension after the year 2024 will depend on the life expectancy. From 2024 the AOW-age will be predetermined five years in advance.
Table 5: Retirement age. (Source: Rijksoverheid, 2013).
Starting in 2019, the pension age will again increase in small steps until 2023 to 67 years. As of 2024 the age to receive AOW will depend on life expectancy (Wilkens, Knoef, Been, Gielen and van Vuuren, 2013). However, it will still be an option to retire at the age of 65. This group will have the possibility to receive an AOW pension, at a net discount of 6.5% per year that they retire earlier than 65. The option to retire early will be used selectively, people with a longer life expectancy will choose to work longer, while people with a shorter life expectancy will choose to retire early (van Vuuren, 2011). However, an increase of 6.5% at delaying the AOW-pension is actuarial low, therefore van Vuuren (2011), expects that the incentive to postpone the AOW-pension will be uncommon. Still, the base of this proposal is to stimulate working for a longer period. As you can see in graph 11, the real retirement age was about 61 before 2006, after 2006 the real retirement age increased
gradually. This increase in the real retirement age is due to the introduction of legislative amendments and regulations aimed at restricting early retirement in 2006 (CBS, 2012).
58 60 62 64 66 A ge
Age of retirement
Graph 11: Real retirement age (Source: CBS Statline, 29-01-2014)
The proposal is created such that working longer will be more financially attractive (Kamerstukken II, 2011/12). According to the CPB (2011), the rise in pension age to 67 years will yield an increase of 0.7% of GDP. This rise will therefore contribute to more sustainable public finances.
4.2 Other options to make the Dutch public pension system more robust
The question is now, can we make the Dutch public pension system even more robust? What are the pros and cons of the following options? First, I discuss the idea of abolishing the AOW, followed by a discussion about raising the pension age even higher, thereafter a discussion about making the AOW means-related and at last making the AOW-pensions depending on the number of working years. These four options will be evaluated on the following eight points; Implementation, individual savings, insurance, labor participation, purchasing power, public expenditures, moral hazard and solidarity.
The first point is implementation, at this point I discuss the possibility of implementing the proposal. Is it possible to implement the proposal at this moment or are there any conditions attached which make it impossible to implement the proposal immediately. For example, in some cases in order to implement a proposal the government needs to possess information about people’s income, capital, savings etc. If they do not possess that information the proposal, in which this information is needed, cannot be implemented at this moment. Therefore, if there are any conditions, I discuss them and decide whether it is possible to implement to proposal in the future or if it is not possible at all. The second point is individual savings; what happens with the individual savings if the proposal is implemented according to economic theory? Most economic analyses of individual savings behavior
use a version of the life-cycle theory of individual savings (Modigliani, 1966). The life-cycle theory assumes that people decide how much to consume and to save looking at present and future resources and present and future needs (Modigliani, 1966; Diamond & Hausman, 1983). According to Feldstein (1974), the most obvious implication of the life-cycle theory is that social security, by providing income during retirement, reduces the amount of savings during the working years. Feldstein (1974) describes it more specifically, if the combination of social security tax and benefits has no net income effect, that is, if the individual’s lifetime budget constraint is unchanged, savings will be reduced by just enough to leave consumption during retirement unchanged. Furthermore, Feldstein (1974) argues that, for middle- and low-income households, social security is a complete substitute for a substantial rate of private savings. Based on these assumptions, I will discuss the implications of the proposal on the individual savings.
Also insurance is an important aspect to take into account. Does the proposal have any influence on the degree in which a person is insured? The AOW-pension scheme can be seen as a kind of an insurance against poverty in the retirement period and to dire shocks during the working life, for example injuries which make it impossible to continue to work which will decrease someone’s income and which makes it impossible to save for retirement. According to Draper, et al (2013), individuals, without insurance against longevity, have to save not only for the expected retirement period but also for the possibility of living longer than expected. However, Draper, et al (2013), claims that the consumption plan will be mainly driven by life expectancy, i.e. the precautionary savings for longevity are relatively small because individuals only have utility of consumption when they are alive. This means that when they grow older than the life expectancy, consumption has to fall. This will lead to a declining consumption profile in the retirement period. Draper, et al (2013), show that when there are no means of insurance against both income and mortality risk it leads to old age poverty. Based on these assumptions I will discuss the implications of the proposal on the degree to which a person is insured.
Next is labor participation; what happens to the labor participation rate if one of the proposals is implemented? Do people have more or less incentives to participate? Borjas (2010), claims that the decision of whether to work or not, boils down to a simple question: is the rate at which leisure can be traded for additional consumption, attractive enough to persuade people into entering the labor market? Furthermore, according to Hausman (1980), labor is supplied and leisure is demanded. Since leisure is demanded, people will only enter the labor market if the reward is big enough. Based on these assumptions, I will discuss the implications of the implementation of a proposal on the rate of labor participation.
Furthermore, purchasing power; what happens to purchasing power if one of the proposals is implemented? Purchasing power indicates how many goods one can buy with his or her disposable income. Therefore, I will look at the disposable income of the retirees and what will happen to their purchasing power.
The next point is public expenditures; will the proposal decrease the total expenditures on the AOW-pension made by the government? As shown in the model (paragraph 3.3) and in graph 10, the total expenditures on AOW-pensions is increasing rapidly. One of the major causes of this increase is the rapid growth of retirees, as shown in this model, a one percent increase in the number of retirees leads to an increase of 1.68% in total expenditures on AOW-pensions. Therefore, solutions need to be found in order to make the Dutch pension system more robust.
Second to last; moral hazard. Will the behavior of employees change if one of the proposals is implemented? According to Rosen & Gayer (2010), moral hazard occurs when people obtain insurance against an adverse outcome which leads to changes in behavior that increases the likelihood of the outcome. The existence of moral hazard introduces a fundamental tension into the design of insurance companies. For example, the more that an insurance plan, or pension plan, smoothes risks against poverty in the retirement period or against dire shocks in the working period, the more it leads to inefficient overuse through an increase in risky behavior like decreasing individual savings or lowering labor supply (Rosen & Gayer, 2010). Based on these assumptions, I will discuss the implications of the implementation of a proposal on moral hazard.
The final point is solidarity. Does the proposal have any effect on the solidarity between people? Since the AOW is a PAYG-system, it implies a solidarity between young and old people. Furthermore, a pension implies solidarity between people who live short to people who live long. At the start of the AOW-pension this was not an issue, at that time the gap in life expectancy between social economic cohorts was small. In the last decennia the gap in life expectancy of social economic cohort has increased (Sap, Nijssen & Schipper, 2008). Therefore, I will discuss the implications of the implementation of a proposal on solidarity.
4.2.1 What if we abolish the AOW?
It seems to be the easiest solution, abolishing the AOW-pension scheme. In that case, the government does not need to pay any pensions to elderly, and they will save on their public expenditures. Since the Dutch public pension system consists of three pillars, there are more possibilities in order to save for a pension.
Though, abolishing the AOW-pension scheme comes with many problems. The AOW-pension scheme is a PAYG-system, therefore implementing this proposal disadvantages an entire generation. That generation has paid for their elderly, but now when they are supposed to receive AOW, the younger generation is no longer paying for it. If, for example, you abolish the AOW-pensions tomorrow, the government still needs to pay AOW-pensions to the current retirees because these people already paid a contribution in order to receive an AOW-pension. Also, the people who are not retired yet, but already have paid the AOW-contribution, still need to receive AOW-pensions. Therefore implementing this proposal, abolishing the AOW-pensions, requires a long transition period.
Another important problem is the fact that there are still people who only receive AOW for their pension. They did not had the opportunity to save for their pension in the second or third pillar. The AOW-pension can be seen as an insurance against income risk in the retirement period and against dire shocks during the working period, abolishing AOW-pensions will also abolish this insurance. Draper, et al (2013) show that when there are no means of insurance against both income and mortality risk it will lead to old age poverty. Therefore, if the government would decide to abolish the AOW-pension, it may lead to old age poverty.
According to Draper, et al (2013), individuals without an insurance against longevity have to save not only for the expected retirement period, but also for the possibility of living longer than expected. Notice also, that according to Feldstein (1974), the most obvious implication of the life-cycle theory is that social security, providing income during retirement, reduces the amount of individual savings during the working years. Therefore, if the AOW-pension is abolished, it will increase the individual savings.
Furthermore, what would be the effect of this proposal on labor participation? Borjas (2010) claims that the decision whether to work or not depends on whether it is attractive enough to persuade people into entering the labor market. Since the AOW-pension is abolished, the total income is decreased. Therefore, due to this negative income effect, it is more attractive to enter the labor market in order to save for retirement.
Graph 12 shows that people above the age of 65 have a lower purchasing power compared to a household without a member above the age of 65. Note that the 65-plus cohort has on average a relatively high standardized income, if you look at the retirees who only receive an AOW-pension their purchasing power is even lower. If you were to abolish the AOW-pension scheme, the purchasing power of the elderly in the society will probably decrease to an even lower level. Especially since the group of elderly is increasing, as seen before (graph 2), it will have an enormous impact on the total purchasing power.
Graph 12: Dynamic Purchasing power (Source: CBS Statline, 4-02-2014)
Still, Van der Lecq (2010) argues that an option to decrease the financial problem of the Dutch government is to abolish the AOW-pension scheme. She also argues that many people can build the AOW part within their supplementary pension, but people who are not able to do so, should receive a senior financial support from the government in order receive a minimum income (van der Lecq, 2010). With this reform, the state pension is closer to how the founder of the AOW-pension scheme ever meant it, a minimum income for every elderly who would otherwise live in poverty. Van der Lecq (2010), proposes to abolish the AOW-pension scheme, lowering the franchise in the second pillar to zero and expanding the senior assistance. The franchise in the second pillar is that part of the salary on which no pension is build, because people will receive the AOW-pension instead. If people do not need to pay a premium in order to receive AOW when they are retired, they are able to build a pension in the second pillar over a larger portion of their assets. Another important factor that is needed in order to abolish the AOW-pension scheme according van der Lecq (2010), is to increase the senior assistance. As mentioned before, this is needed in order to support the cohort of retirees who were not able to accumulate their own socially minimum income for their retirement. This way the senior assistance would guarantee a basic income for those who are beneath the socially minimum income, the remaining cohort would build up their own pension in the second or the third pillar (van der Lecq, 2010). The obvious advantage of this proposal is that the government no longer needs to pay for the AOW-pensions to every person above the age of 65, and even though the fact that premiums for AOW-pensions are no longer paid in this scenario, it has a positive effect on the state budget (van der Lecq, 2010). However, the senior assistance needs to be paid, this must be paid from general taxes since the premiums for AOW-pensions are not longer paid.
Simultaneously, which also van der Lecq (2010) acknowledged, there is the moral risk. Employees who are not obliged via their employer to participate in a collective pension, could not participate
-2 0 2 4 6 8 %
Dynamic Purchasing Power
deliberately in order to take advantage of the senior assistance. Not participating might be more attractive for the cohort that builds a pension which will not be much higher than the retirement assistance (van der Lecq, 2010). Besides, there will also always be a group of people who are not aware of the fact that they should participate in a collective pension. This cohort also has to make use of the senior assistance when they are retired.
The last point that should be discussed is solidarity. Abolishing the AOW-pension will have a negative effect on solidarity. Since the AOW-pension is a PAYG-system it implies a solidarity between the young cohort to the old cohort (Sap, et al, 2008). Abolishing this system will ensure that this solidarity no longer exists.
Implementation Individual Savings Insurance Labor participation Purchasing Power Public expenditures Moral hazard Solidarity - + - + - + - -
Table 6: Effects of abolishing AOW-pension scheme (+ = positive effect, - = negative effect, 0 = no effect).
In table 6, you can see the effects of abolishing the AOW-pension scheme on different important aspects as discussed in this paragraph. The most important issue is whether this proposal can be implemented at this moment. As mentioned before, implementing this proposal will take a few decades and it will affect many people. Therefore, I already conclude that this proposal is not the best solution in order to make the Dutch public pension system more robust.
4.2.2 What if we raise the pension age even higher?
Another option to relieve the burden on public expenditures is to increase the pension age. Since the Dutch government is already increasing the pension age, I examine whether it is possible to increase the pension age even higher. As you can see in the model created in paragraph 3.3, the number of 65-plus has a significant effect on the total spending on AOW. If the Dutch government increases the pension age, the group of people receiving an AOW-pension will decrease, this should imply a decrease on the expenditures on AOW. Raising the pension age would also encourage people to retire at a later age, as you can see in graph 11 this is currently happening due to the regulations aimed at restricting early retirement and the introduction of legislative amendments (CBS, 2012). Therefore, a rise in the pension age is a good option to create this effect. A raising life expectancy and better living conditions are good arguments to implement this proposal (Goudswaard, 2002; Bohn, 2005; etc.).
Graph 13: Life expectancy from 65+ years (Source: CBS Statline, 10-12-2013).
As shown in graph 13, when the AOW was implemented in 1957, the expected remaining life time of a person aged 65 was approximately 15 years, this is increased to about 20 years. In addition, when you look at the forecast of the life expectancy from retirees (at age 65) in graph 14, it is expected to continue to increase. Therefore, raising the pension age is a valid argument in order to relieve the burden on the public expenditures due to the aging problem.
However, implementing this proposal should be done step-by-step because people need the chance to adjust to the new situation. Therefore, it may take a few years before this proposal will be fully implemented. However, it is possible to start implementing this proposal currently, since the government is already increasing the pension age.
Graph 14: Forecast life expectancy from 65+ years (Source: CBS Statline, 30-01-2014).
Not only the total amount of retirees will decrease due to an increase in the pension age, also important to notice is that the participation rate of elderly in the labor market will increase
0 5 10 15 20 25 1951 till 1956 1961 till 1966 1971 till 1976 1981 till 1986 1991 till 1996 2001 till 2006 2006 till 2011 Ye ar s
Life Expectancy from 65+ years
Men Women Average
10 15 20 25 30 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2052 2054 2056 2058 2060 Ye ar s
Forecast Life Expectancy from 65+ years
(Goudswaard, 2011). Since people do not longer receive an AOW-pension when they have reached the age of 65, it is more attractive to continue working (Hausman, 1980).
Since the AOW is a PAYG-system, a larger working cohort implies lower expenditures from the government paid from general resources. According to the CPB (2011) increasing the pension age to 67 years yields an improvement of 0.7% of GDP. This means it will contribute to a more sustainable public finance. If the pension age is raised to 68 years, it will even yield an improvement of 1.0% of GDP (CPB, 2011). Note that according to the CPB (2011), not only direct savings were taken into account, but also other effects like a positive participation-effect and the negative effect of an increase in demand in other financial aid like WW (‘werkloosheidswet’, Law of unemployment), WIA (‘wet werk en inkomen naar arbeidsvermogen’, Law of work and income according to labor capacity) and WWB (‘wet werk en bijstand’, Work and Social Assistance Act). So, raising the pension age even higher than 67 years seems to be even more profitable.
Also, since the AOW-pension can be seen as an insurance against poverty in the retirement period, increasing the pension age will have no effect on this insurance. Since the insurance level will not change due to the increase in pension age, people will have no incentive to change their behavior that would increase the likelihood of the outcome (Draper, et al, 2013). This means that increasing the pension age also does not have an effect on moral hazard.
There are several problems with a rise in pension age. For example, van der Lecq (2010), argues that increasing the pension age is not a solution to the aging problem. The upward movement of the life expectancy seems to be a trend rather than an exception, as can been seen in graphs 12 and 13. Van der Lecq (2010), argues that raising the pension age seems to be outdated at the time that the adaptation is provided. However, the Dutch government already implemented the incentive to increase the pension age starting in 2024; from then on the age to receive AOW-pension will depend on the life expectancy, as can been seen in table 5.
Furthermore, solidarity could be a problem. As can be seen in graph 16, if the pension age will be increased, low-skilled people have to hand in a larger proportion of their retirement compared to high-skilled people (Sap, et al, 2008). Another problem with a rise in the pension age could be that not everybody is able to continue practicing their job after a certain number of years. For instance, physically demanding jobs (e.g. construction workers, firefighters, etc.) are in general, too heavy to maintain when the workers get old (Goudswaard, 2011). On the other hand, mentally demanding jobs (e.g. lawyers, administrative workers, etc.) are generally, easier to maintain. Therefore, the possibility of continuing such a job while employees are becoming older is much higher. In most cases, blue collar workers have had a lower education compared to white collar workers.
Graph 15: Participation rate elderly between 55-65 years (Source: CBS Statline, 30-01-2014).
As can be seen in graph 15, there is already a gap in the participation rate between high and low educated elderly workers. Raising the retirement age will probably increase this gap. Notice, in graph 15, that changes in the regulations of retirement do have a positive effect on the participation rate. As shown before in graph 11, the average retirement age was increasing, this can also be seen in graph 15. Starting from 2006, elderly between the age of 55 and 65 are increasingly participating in the labor market. Higher participation by the elderly has a positive effect on the public finances (van Ewijk, 2011). But as can been seen in graph 15, it will also lead to inequality. According to Goudswaard (2011), in the case of physically demanding jobs, the government should not target early retirement, but on the timely switch (if necessary with the help of training) for positions that are less harmful. However, Falba, Gallo and Sindelar (2009) found the mental health of older employees will be negatively affected if it departs from a sooner or later expected retirement age. Falba, et al (2009) assume that less mentally healthy workers are less productive. Therefore reforming the pension age could have a negative influence on the productivity of elderly. Montizaan (2011), has shown empirically that changes in the retirement age indeed lead to an increased chance on depression, this effect does not change over time. However, offering training sessions to employees will increase the job satisfaction, which can positively affect motivation in the workplace and therefore increasing productivity (Montizaan, 2011). Cuelenaere and Chotkowski (2008), came to the conclusion that many elderly gradually want to reduce their work and eventually want to work part-time.
According to the life-cycle theory, people decide how much to consume and to save looking at present and future resources and present and future needs (Diamond & Hausman, 1983). Because people already have taken in mind that they will retire at a certain age, the lifetime budget constraint needs to be adjusted when the retirement age will increase. Therefore, at the time of
0 20 40 60 80 %
Participation rate elderly between 55 and 65 years
implementation of this proposal, the savings of individuals will increase in order to leave consumption during retirement unchanged (Feldstein, 1974).
At last what will happen to the purchasing power of the retirees? Since the AOW-pension is simply postponed, assuming constant income, the disposable income will be unchanged and therefore this proposal shall not have an effect on the purchasing power of elderly.
Implementation Individual Savings Insurance Labor participation Purchasing Power Public expenditures Moral hazard Solidarity + + 0 + 0 + 0 -
Table 7: Effects of an increase in the pension age (+ = positive effect, - = negative effect, 0 = no effect).
In table 7, shows the effects of increasing the pension age on the eight important aspects discussed in this paragraph. Implementing this proposal is currently possible, which makes this a very attractive option. Also, since this proposal has positive effects on individual savings, labor participation and most important on the public expenditures, I conclude that increasing the pension age is a valid argument to relieve the burden on the public expenditures.
4.2.3 What if we make the AOW income related?
The third option I discuss in order to make the total expenditures on AOW-pensions decrease is to make them income related. In most of the countries in the European Union, the first pillar is (partly) means-related (Goudswaard, 2002). It means that people who earned a low income during their working life will receive a higher AOW-pension compared to people who earned a higher income during their working life. Because the Dutch pension system includes three pillars, people have different possibilities to ensure they will have a good pension during their retirement in more ways than just the AOW-pension. Because of these pillars people can save for their pensions while they are still working, they can save a part of their income for when they are retired.
Sap, et al (2008) from a left winged party in the Netherlands, argue that, since the AOW-pension is a PAYG social security system characterized by life-time annuities and proportional contributions, it implies a solidarity of people who live short to people who live long. At the start of the AOW-pension this was not an issue, at that time the gap in life expectancy between social economic cohorts was small, as can been seen in graph 13. In the last decennia the gap in life expectancy of social economic cohort has increased (Sap, et al, 2008). As shown in graph 16, low-skilled men live about five years shorter compared to high-skilled men, for women this difference is over two and a half years. An even bigger difference is the gap of low- and skilled workers in good health, men who are
high-skilled live about sixteen years longer compared to the low-high-skilled, for women this difference is about fourteen years.
Graph 16: Difference in life expectancy to education and gender (Source: Sap, et al, 2008)
This means that low-skilled receive a limited amount of AOW compared to high-skilled (Sap, et al, 2008). Though, low-skilled have paid longer for the AOW-pension compared to high-skilled because in general, low-skilled worker start to work earlier in their life compared to high-skilled workers (Sap, et al, 2008). Therefore, this proposal has a positive effect on solidarity.
According to Hollanders & Verbon (2010), the AOW is created to guarantee elderly a minimum income level, so when people are able to provide the minimum income level themselves they do not necessarily need to receive more to have a comfortable retirement. Hollanders & Verbon (2010), argue that people with a supplementary pension should receive a lower state pension than those who are entirely dependent on the AOW-pension scheme. Because people with a high income are able to save more during their working life comparing to people with a low income, you might argue that people with a high income do not necessarily need a full AOW-pension from the government because they will be able to provide for an minimum income level themselves. In that way it will be economically justified to give high-income earners less AOW and give low-income earners receive a higher AOW (Hollanders & Verbon, 2010).
However, since retirees will be cut in their public pension, the purchasing power will decrease (CPB, 2012). This has to do with the fact that due to this proposal, people need to increase their savings in
order to receive a pension during their retirement, therefore their disposable income will decrease which will lead to a lower purchasing power.
Though there are more negative effects about this proposal. For example, having a means-tested pension, people are not fully insured against poverty in the retirement period and against dire shocks in the working period. People with a high-income profile are less insured against poverty in the retirement period compared to people with a low-income profile, since the amount of AOW depends on someone’s income.
Furthermore, this proposal can easily lead to a moral hazard problem. People will have less incentive to work harder and earn more if that means that they will receive a limited amount of AOW. Even more important is that this will be the cohort where much participation can be achieved: the low-skilled men and women(Sap, et al, 2008). Also, low-low-skilled or poor individuals will reduce their individual savings in order to receive a higher AOW-pension (Draper, Nibbelink and Uhde, 2013). This means that if we make the AOW-pensions income dependent, it will lead to labor- and savings distortions. This proposal will also lead to more implicit taxes since retirees with a high income do not receive a full AOW-pension even though they have paid the premium. Implicit tax is the part of the AOW-premium where there is no value in the form of a higher pension in return (van Ewijk, 2005).
At last, is it possible to implement this proposal? Implementing this means-tested pension can take a long time because the Dutch tax authorities only safe data about people’s income for the last five years. Therefore, the government does not own the information needed in order to implement this proposal currently, which means that this proposal cannot be implemented at this moment.
Implementation Individual Savings Insurance Labor participation Purchasing Power Public expenditures Moral hazard Solidarity - - - - - + - +
Table 8: Effects of a means-dependent AOW-pension scheme (+ = positive effect, - = negative effect, 0 = no effect). In table 8, you can see the effects of abolishing the AOW-pension scheme on different important aspects as discussed in this paragraph. The most important issue is whether this proposal can be implemented at this moment. As mentioned before, implementing this proposal will take a few decades since the Dutch tax authorities do not possess sufficient information in order to implement this proposal. Therefore I can already conclude that this proposal is not the best solution in order to make the Dutch public pension system more robust. In fact, since the only two positive effect of this proposal is on public expenditures and on solidarity, I conclude that it is never the best solution to implement this system even if sufficient information is collected.
4.2.4 What if we make AOW dependent on the number of working years?
The last proposal I discuss is the possibility to set a minimum amount of working years in order to receive an AOW-pension. As seen in paragraph 4.2.2, the major issue of increasing the pension age is that it is not possible for everyone to continue practicing their job after a certain amount of years (van Ewijk, 2011). As can been seen in graph 15, there is already a large gap in the participation rate between high- and low-educated elderly workers. Jobs could be too physically demanding when the workers get older, while other mentally demanding jobs, are in general easier to maintain. Sap, et al (2008), argue that in general most of these workers have had a lower education compared to white collar workers.
Furthermore, in paragraph 4.2.3, the main advantage of making the AOW-pension income dependent was that because the AOW-pension is a PAYG-social security system which is characterized by life-time annuities and proportional contributions, it implies a solidarity between people who live short to people who live long. In the last decennia the gap in life expectancy of social economic cohort has increased (Sap, et al, 2008). As was shown in graph 16, lower educated men live about five years shorter compared to high educated men, the difference for women is over two and a half year. The differences between low- and high-educated people is even higher if you look at the number of years in good health. According to Sap, et al (2008), low-educated workers start earlier in their live to work compared to high-educated workers. This is due to the fact that lower education does not take as long as higher education. Therefore, if we make the AOW-pension dependent of the number of years that someone has worked it seems to be a valid requirement. Bovenberg, Mackenbach and Mehlkopf (2006), argue to link the retirement age to the life expectancy by profession. Bovenberg, et al (2006), argue that a differentiated retirement age is, besides it is more fair, more effective. A general increase of the retirement age would, for the cohort with low healthy life expectancy, only result in a shift of the costs of pensions to costs for disability (Bovenberg, et al, 2006).
According to Sap, et al (2008), an employment-related AOW provides the right incentives for a higher participation rate because it is better connected with the actual capabilities of people to work longer. Therefore, implementing this proposal will have a positive effect on labor participation.
Furthermore, there will probably be no effect on the purchasing power of elderly, since every person still receives an AOW-pension and is able to save in the second and the third pillar of the pension system. The disposable income of elderly will not change which means that nothing happens with the purchasing power.
