Dissertation

Table 3.22: Decomposing the labor supply responses under the efficiency profile from Hansen (1993).

(a) The intensive margin: average weekly hours spent in market work between ages 25-55.

Experiment ϕ = 0.18 ϕ = 0.385 ϕ = 0.59 ϕ = 0.795 ϕ = 1 Baseline 29.3 33.79 35.14 35.8 36.18 1 28.82 33.73 35.21 35.94 36.36

2 28.78 33.93 35.49 36.25 36.8

3 28.89 34.11 35.7 36.47 36.92

(b) The extensive margin: actual retirement age. Experiment ϕ = 0.18 ϕ = 0.385 ϕ = 0.59 ϕ = 0.795 ϕ = 1

Baseline 59.52 61.32 61.9 62.18 62.35 1 66.23 67.16 67.41 67.52 67.59 2 67.09 67.88 68.1 68.2 68.26 3 67.82 68.51 68.7 68.8 68.85

77

Table 3.23: Equilibrium social security benefits with the optimal tax response under Hansen (1993).

Experiment ϕ = 0.18 ϕ = 0.385 ϕ = 0.59 ϕ = 0.795 ϕ = 1 Baseline 0.034 0.0468 0.0595 0.0723 0.085

1 0.0287 0.0395 0.0503 0.061 0.0718 2 0.028 0.0385 0.049 0.0595 0.0701 3 0.0277 0.0381 0.0485 0.059 0.0694

Finally, I report in Table 3.23 social security benefits in post-population aging steady

states with the welfare-maximizing tax responses. The table reveals a similar pattern, in

the sense that benefits decline by 15.5, 17.5 and 18.4% under the three experiments with

the optimal tax response. Also, these decline are significantly smaller than what would have

occurred if the tax rate were held fixed under population aging: declines of 28, 35.5 and

42%, respectively.

Sensitivity analysis with respect to the parameters in the age-dependent component of

household efficiency shows that the quantitative predictions of the model are fairly robust.

The welfare-maximizing social security tax rates under population aging are roughly 2-5

percentage points higher than the baseline value even with the efficiency profile estimated

from Hansen (1993).

It is worth noting that all the simulation results so far have been based on the as-

sumption that the age at which households start receiving social security benefits (T_r) will

remain unchanged under population aging. However, the Social Security Administration

has been steadily increasing the full retirement age in the U.S. over the past several decades.

Currently, the full retirement age for an individual depends on when he/she was born (see

Table 3.24). In their applied general equilibrium study on the projected U.S. demographics

and social security, De Nardi et al. (1999) consider a fiscal response in which the full re-

tirement age is postponed twice by 2 years each time in 2032 and 2036, eventually raising

it to 69. What impact would such a postponement in full retirement eligibility have on

the quantitative predictions of the current model? To examine this, I design a new set of

computational experiments in which I increase the age at which households start receiving