Based on the gov't mortality tables, I assume the NPV of SS is the same whether you take it at 62, FRA, or 70. However, I wonder which of these scenarios actually works out more favorably for the gov't or does it indeed correlate perfectly with the mortality tables?
Well, by the tables currently in use by SSA, they expect some folks to live well beyond the century mark (Male age: 111; Female age: 113):
Actuarial Life Table
I would go with the general assumption that regardless of when you start SS (62, FRA, 70) it might be a "wash" for an indivudial. This can be tested by plugging in the various age/amounts into your retirement income forecast tool of choice.
However, for married folks this "automatic assumption" goes out the window, based upon their ages, SS credits, and desires (such as my waiting till age 70 for the benefit of DW, while at the same time tapping into her benefit when she submits her claim at FRA).
It's truly a mesurement and a decision based upon many factors in a married (or even previously married, at least 10 years) situation.
The government also has to consider the "tails" in the situation. Those who may collect benefits due to their age and the death of one/both parents.
And then there the question of SSD (Social Security Disability - don't confuse the term with SSI which is a completely different program that pays SS to disabled folks who never wo*ked).
Not to be political (sorry), but the "government" dosen't make out one way or another, other than to measure the "health" of the system and look/suggest ways of adjusting inflow/outflows, IMHO.