9.6  Other approaches to partially prefunding the NZ Superannuation Fund (7.3)

This section shows that the current funding method for the NZ Superannuation Fund is insufficient to control the liabilities of the scheme relative to GDP. If the unfunded liability is treated as a government liability, the government net savings position will deteriorate under the current funding method.

If the government were to adopt a SAYGO approach to funding NZ Superannuation, the funding rate for new entrants to the workforce would be quite low, between 1% and 2% of GDP (using Treasury's NZSF funding model). However, there would be a large unfunded liability (UFL) for existing members of the workforce.

The cost of amortising the unfunded liability over any reasonable period is high. However, one approach that is often used for pension schemes is to contribute to the fund an amount equal to the valuation rate of interest on the unfunded liability so that the unfunded liability stays constant in nominal terms. Keeping the NZ Superannuation unfunded liability constant in nominal terms is also expensive, but a more realistic option would be to keep it constant as a share of GDP. Indeed, if we don't, we are incurring an increasing liability that must contribute to the same problems as a high NFL. (This is why the projected contribution rate in Figure 9.14 is increasing: the total contribution rate is less than the funding rate for new entrants plus the amount required to keep the unfunded liability constant as a share of GDP.)

Figure 9.14: Contribution rate to NZSF under current funding approach is rising as a GDP share[56]

Source:  SWG calculations

Figures 9.14, 9.15 and 9.16 compare the current funding method with a funding approach that holds the UFL constant as a percentage of GDP. Because the fund is taxed, the net cost to the government is the contribution rate less tax receipts from the fund, shown by the dashed lines in Figure 9.14. Note that the contribution rate required to hold the UFL constant relative to GDP does not increase as fast as the current method (after the short-term projection effect).

Figure 9.15 shows that the under the current funding scheme the UFL increases over time: the contributions are insufficient both to fund new entrants and to keep the UFL constant as a share of GDP.

Figure 9.15: Under current funding, the contribution rate is not enough to fund new entrants and keep UFL a flat share of GDP

Source:  SWG calculations

Because the contributions are higher when the UFL is stabilised, the fund grows more quickly, as shown in Figure 9.16.

Figure 9.16: For UFL to be a stable share by 2050, the Fund size needs to double compared with the current approach

Source:  SWG calculations

Figures 9.17, 9.18 and 9.19 show the growth of the Fund, if it was not taxed. Contributions are lower, and the Fund grows much larger. Note that if investment earnings are not taxed, the UFL under the current funding scheme grows much larger than under the current arrangement.

Figure 9.17: With earnings not taxed, contributions are smaller to hold UFL as a fixed GDP share

Source:  SWG calculations

Figure 9.18: With earnings not taxed, the UFL as a GDP share is about a third smaller than in the taxed case

Source:  SWG calculations

Figure 9.19: With earnings not taxed, by 2070 the Fund more than doubles compared with taxed case

Source:  SWG calculations

This analysis shows that the current funding method is insufficient to control the liabilities of the scheme relative to GDP. Treating the UFL as a government liability (as would be the case for a company under current accounting conventions), the government's net savings position will deteriorate under the current funding method.

This position could be rectified by:

1. Making the NZ Superannuation Fund tax exempt. Stabilising the UFL relative to GDP would require a total contribution rate not much higher than the current rate. The net revenue effect to the government would be changed only by timing.
2. Perhaps introducing a dedicated social security tax (with an offset to ordinary income tax) equal to the new entrant fund rate. This rate would be about 4.2% of the average wage for all employed persons (this has been adjusted for those not in the work force) using Treasury assumptions, equivalent to 2.4% of GDP. The funding rate for new entrants would be 2% of the average wage, equivalent to 1.1% of GDP. The remaining contributions would be paid by the government as at present. A social security tax would create a discipline of continuing to save regardless of the fiscal position, and current anomalies in the benefit formula could more easily be repaired.