You could give the user the option of specifying their precise SS earnings for past years, and projection for the future. This old tool could serve as a prototype for how far you might go with it: https://www.ssa.gov/OACT/anypia/download.html Once you have the monthly payment estimate I’d suggest having a % reduction factor to account for predictions of insolvency. details: https://www.ssa.gov/policy/docs/ssb/v70n3/v70n3p111.html More recent discussion of insolvency puts it earlier, 2034 at 78% of benefits.

Two other open source tools that may be useful are:

https://github.com/Gregable/social-security-tools

For calculation of PIA.

and

https://github.com/MikePiper/open-social-security

for deciding when to claim social security benefits. Since ProjectionLab has the amount of savings, it can factor that into the social security claiming decision.

SSA.gov will now tell you your expected payment, and let you select $0 for Average Future Annual Salary. If PL users input this single value, you should be able to project out actual benefits based on the plan’s earned income, and selected retirement age.

There are so many unique nuances of SS that any “estimator” would risk being vastly inaccurate. Given the fact that that SSA.gov already does this for each person based on their own personal anticipated Social Security, as does opensocialsecurity.com, I don’t think PL needs to add it. Yes, it would be cool, but I’d rather they focus on other things elsewhere on this website.

Love this idea

First off, adding this as another very useful and powerful estimator https://ssa.tools/calculator

and then adding the commentary that the SSA.gov estimator is limited more than I think a typical PL user might be willing to accept. Even the SSA.gov tool isn’t equipped to handle the many “special” cases and that’s why they recommend their downloadable PIA tool. And even that isn’t perfect for certain other cases which is why they say the real final answer just can’t be known until you file.

Or said another way, all models are wrong, some are useful and many of the online tools are a little more useful than even the SSA.gov one.

That said, I think it would be better for Kyle to focus on the core functions of PL for now before getting into the game of keeping an SS estimator updated. What he CAN do is offer links to various estimators (or even the boglehead wiki’s estimator links page).

Now, THAT said, it would eventually be nice for PL to do the math based on forward looking employment plan projections especially for people modelling future career changes, going back to school, etc, which can significantly impact SS tax contributions. Even https://ssa.tools/calculator only allows you to set a single value for future earning which might not always be true. But it’s a useful simplification for a free web app. The math itself isn’t even that hard, I’ve done it in excel.

I agree with the bits about not trying to reproduce calculators for a system with so much complexity.

But I did once find myself surprised by a very optimistic projection. The explanation was that I modeled taking Social Security at a lower age, but forgot to hand edit the monthly amount.

It made me wonder about a smaller extension of PL’s Social Security Income “module” to allow input of a PIA number from a favorite calculator and then PL would automatically adjust the monthly amount based on the user-supplied claiming age.

That said, my personal SS situation feels pretty generic. There may be complexity I’m not factoring in with that proposal.

I like this idea but because of all the complexities and possibilities of errors, I wish we could enter the amount for both retiring at age 62 and at age 70. The in between points can then be linearly interpolated. Then we could more easily see what happens if we decide to start Social Security at different ages

You don’t want a linear interpolation between 62 and 70, as it’s not correct. Just have a field for the full retirement age PIA and use the SSA formulas to calculate the benefit at different ages. The correct calculations aren’t that complex.