Page 99 - Invited Paper Session (IPS) - Volume 1

IPS102 Arjan B.
voluntary pensions by age and gender, which we impute in our household
database and balance it proportionally to arrive at the macro total.
With regard to the non-financial assets, we distinguish dwellings, land
underneath dwellings, transfer costs, and company assets of self-employed.
For the first three items we use the Valuation of Immovable Property Act (In
Dutch: WOZ), for the distributional information. This data source is available
as a register, and can be linked to individuals using the unique key. Company
assets are distributed over individuals using the consumption of fixed capital
in accordance with the current accounts.
For the macro estimate of public pension entitlements, the population is
divided in two groups: those who accrue entitlements, and those who receive
a pension benefit. For the latter we have actual benefits from the IIWS, for the
former we assume that entitlements increase up to the average pension
entitlement in that year. In the Netherlands, the first pillar does not depend
on labour history, simply on the years one has been a resident in the country.
In fifty years a full benefit can be accrued, hence we assume a two percent
4
yearly increase . Similar to the approach for the work-related pension
schemes, this is combined with the life expectancy and the foreseen retirement
age, to estimate a NPV of all entitlements. This requires discounting, which is
the most influential parameter in the calculation. Eurostat proposes a 3% real
discount rate. We deviate from this proposal with the main argument that this
would make the work-related and the public pension entitlements
incomparable. Pension funds in the Netherlands are obliged to use the IRTS,
which compares poorly with the Eurostat proposal. First, because the Eurostat
approach treats each entitlement the same, while the IRTS differentiates by
loan term. Second, the 3% real rate seems too positive in recent years. Loan
terms of four years and shorter even had a negative discount rate in the IRTS.
Entitlements with a loan term far in the future have the highest discount rate,
but with a nominal 2.407% in 2016, this was far from the 3% real interest rate.
Apart from the discount rate, we also deviate from the proposed approach for
estimating the public pension entitlements. Two approaches are possible, the
projected benefit obligations approach (PBO), and the accrued benefit
obligations approach (ABO). The PBO approach estimates the entitlements
including future wage increases or indexation of pensions. The ABO approach
does not include these. Eurostat proposes the former approach, but we favour
the latter. Again we argue that ABO best reflects SNA practice, and that it
makes our resulting public pension entitlements comparable with the work-
related pensions.

4 We assume that any benefit-gap, because one may have resided outside the Netherlands for
some time, is included in the average benefit in a year and thus in the average accrued
entitlement in that year.
88 | I S I W S C 2 0 1 9