Kaufmann et al. (2024) calculate the MVPF of a 2006 Dutch pension reform which incentivized people to work longer. Using administrative data covering the universe of the Dutch population and a fuzzy Regression Discontinuity (RD) design, the paper studies the impact of grandparental retirement decisions on family members’ labor supply and child outcomes. Specifically, the paper exploits a Dutch pension reform in 2006 which made early retirement less attractive for people born in 1950 or later. The paper finds that a one-hour increase in grandmothers’ hours worked causes their adult daughters with young children to work 40 minutes less. When examining the reform’s impact on grandchildren’s test scores, the paper finds positive effects on children aged 4-7 who experienced a substitution from grandparental to maternal care. The paper also finds negative effects for children aged 8-12, for whom grandparental childcare was substituted for by formal or no care.

Kaufmann et al. (2024) calculate the MVPF of the 2006 Dutch pension reform, taking into account the response by grandmothers and their daughters. The MVPF is framed as the MVPF of the reduction in pension benefits, yielding a negative net cost as well as a negative willingness to pay.

MVPF = 0.6

The net cost of a policy incentivizing later retirement is the sum of the direct costs (or savings) of the government plus any fiscal externalities as a result of delayed claiming and changed labour supply.

The mechanical net cost of incentivizing later retirement is the change in pension generosity. The government saves €7,013 per person on average through reduced pension payments.

There are several components of the fiscal externality. First, the government saves €1,833 per person on pension payments due to delayed claiming by around 5 months.

In addition, the government collects €2,445 more income taxes per person over five years because grandmothers work more between ages 60 and 64.

On the other hand, the government forgoes additional income taxes of €1,585 per person over five years due to reduced earnings by mothers.

The net cost is then -€7,013 – €1,833 – €2,445 + €1,585 = -€9,706. In other words, the government saved €9,706 per family of an affected grandmother with adult daughter whose youngest child is between age 4 and 12. This savings estimate is conservative as it does not account for the increase in earnings due to an increase in earnings from the sons-in-law of the grandmothers who work longer.

That paper notes that, if they consider the long-run reform effects on mothers’ lifetime income and assume

that government only cares about income tax revenue, the loss in tax revenue due to the drop in maternal labor supply would outweigh the gain in tax revenue from delaying the retirement of the grandmothers if the impact on maternal labor supply lasts for up to eight years. The net cost is still negative, however, aver accounting for the changes in pension payments.

Furthermore, we calculate the MVPF from a perspective of the affected grandmother’s family. If we consider the family of the affected grandfathers, the net costs will be even larger, as grandfathers delay retirement more and the spillover effects on his daughter are smaller.

-€9.7K
Net Cost

Upper Margin
Lower Margin

The willingness to pay by grandmothers for a policy incentivizing later retirement is the sum of what would be the sum of their forgone public pension payments (-€7,013 -€1,833) and their increased earnings from working more (€7,020=€117*12*5), for a total of -€1,826.

The willingness to pay of the adult daughters whose youngest child is between 4 and 12 years for a policy incentivizing later retirement is the reduced earnings of -€3,480 (=€58*12*5).

The total willingness to pay per person is then -€1,826 – €3,480 = -€5,306. If the paper included the willingness to pay of the sons-in-law (who experienced an increase in earnings on average), the absolute value of the willingness to pay would be smaller, but the value would still be negative.

-€5.3K
WTP

Upper Margin
Lower Margin

The estimated MVPF of the 2006 Dutch pension reform—taking into account the response by grandmothers and their daughters—is then (-€5,306/ -€9,706) = 0.55.

The paper highlights the importance of incorporating the spillover effects by noting the MVPF would be smaller (-€1,826/-€11,291)=0.16 if the spillover effects on adult daughters were excluded from the calculation. Although the fiscal gain is smaller when considering the spillover effects, the WTP to avoid this policy also gets larger in absolute terms.

While not included in the main MVPF estimate, the paper finds positive effects on test scores for young girls and boys, and negative effects on older boys. The paper notes that, if these spillover effects on grandchildren were included, the expected impact on government tax revenue would likely be positive (or at least non-negative). This would further reduce the MVPF by increasing the net savings to the government and also increasing the absolute value of the WTP.

0.6
MVPF

Upper Margin
Lower Margin

Kaufmann, Katja M., Yasemin Özdemir, and Han Ye (2024). “Spillover Effects of Old-Age Pension Across Generations: Family Labor Supply and Child Outcomes.” Working Paper.

https://www.dropbox.com/scl/fi/tqmwqbxwkdenyvmj2l7h9/Spillovers_old_age_pension.pdf

- Category
- Social Insurance
- Sub-Category
- Retirement
- Beneficiary Type(s)
- Adults, Seniors (65+)
- Country of Implementation
- Netherlands
- Year of Implementation
- 2006
- Empirical Method
- Regression Discontinuity
- Research Type
- Primary
- Peer Reviewed
- No
- MVPF Publication Link
- www.dropbox.com/scl/fi/tqmwqbxwkdenyvmj2l7h9/Spillovers_old_age_pension.pdf