"Open Social Security" calculator: feature requests, bug reports, etc

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 1:23 pm

OK, the new behavior is as follows:

If an input on the primary form is changed (e.g., discount rate or mortality table), clicking the "submit" button in the "alternative date" form will:
1) Run the original calculation again -- getting a new recommended strategy, updating the corresponding PV, and updating the corresponding table, and
2) Calculate the PV using the alternative dates selected by the user -- and update the corresponding table.

If no inputs on the primary form have been changed (i.e., the only thing the user is experimenting with are the alternative dates), then #1 above is skipped and only #2 happens (so the user doesn't have to wait while unnecessary math is done).
Mike Piper, author/blogger

User avatar
FiveK
Posts: 5613
Joined: Sun Mar 16, 2014 2:43 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by FiveK » Mon Aug 13, 2018 1:39 pm

ObliviousInvestor wrote:
Mon Aug 13, 2018 1:23 pm
OK, the new behavior is as follows:....
A+ for responsiveness!

Not that you don't already get high marks for the work itself. :)

In short - thanks!

User avatar
One Ping
Posts: 488
Joined: Thu Sep 24, 2015 4:53 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by One Ping » Mon Aug 13, 2018 3:18 pm

Good work, Mike.

Next challenge. :happy

The table of "Year-by-Year Benefit Amounts" shows, for a given year, the benefit "claimed" amount, not the benefit "paid" amount.

What do I mean by that?

The easiest way understand this is to think about a person who claims their benefit starting in December 2011. Their first benefit is actually paid in January 2012. For someone who claims in December 2011, but whose first payment is in January 2012, the table shows a benefit in December 2011. While maybe this is technically true in terms of when benefits were claimed, that benefit is paid and therefore taxed in 2012, not 2011. My belief is that most people are likely to think of the amounts shown under the annual benefits and total columns as being (potentially) taxable in the year for the row shown. For at least ~8⅓% (1 out of 12) this is likely not true.

Is this a big deal to change? It would certainly make it easier on those trying to take into account some tax planning considerations.

One Ping
"Re-verify our range to target ... one ping only."

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 3:25 pm

One Ping wrote:
Mon Aug 13, 2018 3:18 pm
The table of "Year-by-Year Benefit Amounts" shows, for a given year, the benefit "claimed" amount, not the benefit "paid" amount.

What do I mean by that?

The easiest way understand this is to think about a person who claims their benefit starting in December 2011. Their first benefit is actually paid in January 2012. For someone who claims in December 2011, but whose first payment is in January 2012, the table shows a benefit in December 2011. While maybe this is technically true in terms of when benefits were claimed, that benefit is paid and therefore taxed in 2012, not 2011. My belief is that most people are likely to think of the amounts shown under the annual benefits and total columns as being (potentially) taxable in the year for the row shown. For at least ~8⅓% (1 out of 12) this is likely not true.

Is this a big deal to change? It would certainly make it easier on those trying to take into account some tax planning considerations.
At the moment, this is the intended behavior.

It's true that the benefit for a given month is paid in the following month. And right now the table is based on what your benefit for each year would be.

I understand that this might throw off somebody trying to do very specific tax planning based on the table. But I suspect that doing it the other way would cause additional confusion in many cases, as it would make the output not appear to line up with the claiming dates. As always though, I'm open to hearing what other people think here.

As far as how hard it would be to change, hard to say. My best guess it would take a couple days of full-time work. But I'm really not good at guessing. (I expected voluntary suspension functionality to take a couple weeks. It took a couple months. And the opposite sort of thing has happened a few times -- some changes turn out to be much easier than I anticipate.)
Mike Piper, author/blogger

User avatar
One Ping
Posts: 488
Joined: Thu Sep 24, 2015 4:53 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by One Ping » Mon Aug 13, 2018 3:34 pm

ObliviousInvestor wrote:
Mon Aug 13, 2018 3:25 pm
One Ping wrote:
Mon Aug 13, 2018 3:18 pm
The table of "Year-by-Year Benefit Amounts" shows, for a given year, the benefit "claimed" amount, not the benefit "paid" amount.

What do I mean by that?

The easiest way understand this is to think about a person who claims their benefit starting in December 2011. Their first benefit is actually paid in January 2012. For someone who claims in December 2011, but whose first payment is in January 2012, the table shows a benefit in December 2011. While maybe this is technically true in terms of when benefits were claimed, that benefit is paid and therefore taxed in 2012, not 2011. My belief is that most people are likely to think of the amounts shown under the annual benefits and total columns as being (potentially) taxable in the year for the row shown. For at least ~8⅓% (1 out of 12) this is likely not true.

Is this a big deal to change? It would certainly make it easier on those trying to take into account some tax planning considerations.
At the moment, this is the intended behavior.

It's true that the benefit for a given month is paid in the following month. And right now the table is based on what your benefit for each year would be.

I understand that this might throw off somebody trying to do very specific tax planning based on the table. But I suspect that doing it the other way would cause additional confusion in many cases, as it would make the output not appear to line up with the claiming dates. As always though, I'm open to hearing what other people think here.

As far as how hard it would be to change, hard to say. My best guess it would take a couple days of full-time work. But I'm really not good at guessing. (I expected voluntary suspension functionality to take a couple weeks. It took a couple months. And the opposite sort of thing has happened a few times -- some changes turn out to be much easier than I anticipate.)
OK.
It's just a little "copy, paste, adjust" work in excel. Great tool! I love it. :sharebeer
"Re-verify our range to target ... one ping only."

vtMaps
Posts: 400
Joined: Tue Oct 14, 2008 12:05 pm
Location: central Vermont

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by vtMaps » Mon Aug 13, 2018 4:51 pm

My understanding of Delayed Retirement Credits is that they are not credited to your account until January of the year following the year in which they were earned. Thus if I file at full retirement age (FRA) in March of 2019 or in September (FRA + 6 months), wouldn't my monthly payment be the same until January 2020?

The calculator shows my payments increase each month during the calendar year(s) after my FRA.

--vtMaps
The optimist proclaims that we live in the best of all possible worlds; and the pessimist fears this is true. --James Branch Cabell

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 4:59 pm

vtMaps wrote:
Mon Aug 13, 2018 4:51 pm
My understanding of Delayed Retirement Credits is that they are not credited to your account until January of the year following the year in which they were earned. Thus if I file at full retirement age (FRA) in March of 2019 or in September (FRA + 6 months), wouldn't my monthly payment be the same until January 2020?

The calculator shows my payments increase each month during the calendar year(s) after my FRA.
Indeed. This is one of a few simplifications the calculator makes. Here is some discussion from earlier in the thread:
ObliviousInvestor wrote:
Tue Jun 12, 2018 4:23 pm
MrTimewise wrote:
Tue Jun 12, 2018 4:02 pm
Mike,

Just out of curiosity, did you factor into the total benefit computations the initial benefit year's lower monthly payments that can result when claiming between FRA and 70 years. That is, when claiming after FRA but before 70 years, the SSA's uses only the accumulated DRCs as of January of the year of filing. Then, the following January, any DRCs that had accumulated within the year of filing are additionally used to determine total benefits.
No, that's not included in the calculation.

The biggest impact it can have is one year's benefit being underweighted by 6 months x 6 DRCs = 24% of the person's PIA (e.g., a one-time $240 cost for a person with a $1,000 PIA). That's generally going to have an impact less than 0.2% of the overall NPV -- often less than 0.1% if they file in a month other than July.

If this were enough to sway the calculator from suggesting one claiming month to a different claiming month, it would essentially mean nothing other than "these two claiming months are basically as good as each other" -- which you can already see if the NPVs are that close together.
There are other simplifications as well. For example, all benefits are assumed to be received mid-year and discounted annually, rather than month-by-month discounting. In addition, for the purpose of the earnings test, it's assumed that a person goes from working to completely retired on the date they choose for the related input, whereas in reality people often have a more gradual retirement.
Mike Piper, author/blogger

Keith_Simons
Posts: 3
Joined: Sun Aug 12, 2018 11:31 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Keith_Simons » Mon Aug 13, 2018 6:39 pm

ObliviousInvestor wrote:
Mon Aug 13, 2018 6:35 am

I'm having a hard time following your question without more specific information (i.e., actual dates and PIAs).

A spousal benefit is calculated as:

50% of the other person's PIA, minus the greater of your own PIA or your own retirement benefit. (If this amount is below zero, you get no spousal benefit rather than a negative spousal benefit.)

In the event that a spousal benefit is claimed prior to FRA, the above calculation is multiplied by a reduction factor. Specifically, the reduction is 25/36 of 1% for each month early, up to 36 months. For each month in excess of 36 months, the reduction is 5/12 of 1%.

And that spousal benefit is added to your own retirement benefit to arrive at your total monthly benefit.

(Edited for clarity.)
Quite understandable - here is some relevant data, without revealing personal information.
First claimant - Female - FRA March 2018, first claimed April 2014 (age 62) - PIA 920 at FRA
Second claimant - Male - FRA July 2021, entered as intending to work through end 12-2023, with a PIA at FRA of 2700, earnings per mo (flat rate) at 10K per month, not yet filing for benefits.
Other variables standard (Mortality Tables, PV interest rate)

Calculator strategy recommends:
First suspends retirement benefit till 1/2020, at age 67 and 10 months.
Your spouse files for his/her retirement benefit to begin 5/2025, at age 70 and 0 months.
You file for your spousal benefit to begin 5/2025, at age 73 and 2 months.
Yielding annual benefits for first claimant totaling 14137, of which 4560 is the spousal benefit component, the total representing 34.6% of secondary claimant's annual benefit 40768.
SSA calculator indicates that, regardless of first claimant's voluntary suspension, ratio of first claimant's total benefit to second claimant's benefit should be 35.20%.

Using a comparative claiming strategy of 1-2024 for spousal and spouse's claiming dates yields annualized benefits for first claimant that are unchanged at 14137, of which 4560 is the added spousal component. Second claimant's annualized benefit is 37440, making first claimant's total benefit 37.76% of second claimant's annual benefit.

Shouldn't the ratio be identical, regardless of the spousal claiming date? Shouldn't the ratio agree with the 35.20% calculated at SSA?

User avatar
Watty
Posts: 14135
Joined: Wed Oct 10, 2007 3:55 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Watty » Mon Aug 13, 2018 7:04 pm

ObliviousInvestor wrote:
Mon Aug 13, 2018 8:39 am

This is indeed quite a surprise. I am so far unable to replicate such behavior. Could you perhaps provide the inputs you used?
Here it is.

Windows 10, I just did updates yesterday.
Chrome version, Version 68.0.3440.106 (up to date).

Input, only the things that were not defaults.

Married
Male
DOB: 9/15/56
PIA: $2532

Spouse
Female
DOB: 6/15/51
PIA: 1000

Has Already filed for benefits: Yes 7/2013

Click first submit

Result
The strategy that maximizes the total dollars you can be expected to spend over your lifetimes is as follows:

Your spouse suspends his/her retirement benefit today.
Your spouse unsuspends his/her retirement benefit 2/2020, at age 68 and 8 months. (Monthly benefit: $874)
You file for your retirement benefit to begin 9/2026, at age 70 and 0 months.
Your spouse files for his/her spousal benefit to begin 9/2026, at age 75 and 3 months.
The present value of this proposed solution would be $713,001.
Test an alternative claiming strategy:
Me, 1/2022
Spousal benefits 1/2022

Press second submit, results;
The present value of the strategy you selected is $681,812, as compared to a present value of $713,001 from the recommended strategy, a difference of $31,189.
Scroll back up to "Real discount rate", enter 2

Press first submit button again.

Recommended strategy and present value is unchanged.

With a higher discount rate I would expect the present value of the proposed solution to be a lot lower.

User avatar
Watty
Posts: 14135
Joined: Wed Oct 10, 2007 3:55 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Watty » Mon Aug 13, 2018 7:08 pm

On the question of when to update the numbers one thing I would keep in mind is that people will likely print off the results and show it to someone else so that the context of what has been changed will be lost. It the would be confusing to not update or blank-out the second sets of results when the top calculations are changed.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 8:09 pm

Watty wrote:
Mon Aug 13, 2018 7:04 pm
Here it is.
...
Thank you for bringing this up and providing the necessary inputs! (Given a very specific set of inputs -- the ages and the fixed filing date, then running the original function, the alternative date function, and original date function again in that order -- it was erroneously generating an error message that prevented the function from being run the second time.)

It should be fixed now.
Mike Piper, author/blogger

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 8:10 pm

Watty wrote:
Mon Aug 13, 2018 7:08 pm
On the question of when to update the numbers one thing I would keep in mind is that people will likely print off the results and show it to someone else so that the context of what has been changed will be lost. It the would be confusing to not update or blank-out the second sets of results when the top calculations are changed.
This is a great point. Let me ponder over those options for a bit.
Mike Piper, author/blogger

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Aug 13, 2018 8:23 pm

Keith Simons, I will check the math of your example tomorrow and see if I have made an error, or if there's simply a misunderstanding here of some sort.
Mike Piper, author/blogger

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Tue Aug 14, 2018 7:59 am

Keith_Simons wrote:
Mon Aug 13, 2018 6:39 pm
here is some relevant data, without revealing personal information.
First claimant - Female - FRA March 2018, first claimed April 2014 (age 62) - PIA 920 at FRA
Second claimant - Male - FRA July 2021, entered as intending to work through end 12-2023, with a PIA at FRA of 2700, earnings per mo (flat rate) at 10K per month, not yet filing for benefits.
Other variables standard (Mortality Tables, PV interest rate)

Calculator strategy recommends:
First suspends retirement benefit till 1/2020, at age 67 and 10 months.
Your spouse files for his/her retirement benefit to begin 5/2025, at age 70 and 0 months.
You file for your spousal benefit to begin 5/2025, at age 73 and 2 months.
Yielding annual benefits for first claimant totaling 14137, of which 4560 is the spousal benefit component, the total representing 34.6% of secondary claimant's annual benefit 40768.
SSA calculator indicates that, regardless of first claimant's voluntary suspension, ratio of first claimant's total benefit to second claimant's benefit should be 35.20%.

Using a comparative claiming strategy of 1-2024 for spousal and spouse's claiming dates yields annualized benefits for first claimant that are unchanged at 14137, of which 4560 is the added spousal component. Second claimant's annualized benefit is 37440, making first claimant's total benefit 37.76% of second claimant's annual benefit.

Shouldn't the ratio be identical, regardless of the spousal claiming date? Shouldn't the ratio agree with the 35.20% calculated at SSA?
I wasn't familiar with that calculator on the SSA site, so I had to play with it a little. As it turns out, it isn't doing a very thorough analysis. All it does is assume that you get 50% of your spouse's PIA, and then applies any reduction for early entitlement. The reduction for early entitlement to spousal benefits is 25/36 of 1% for each month early, up to 36 months. For each month in excess of 36 months, the reduction is 5/12 of 1%.

So, in this case, if you enter a March 1952 DoB and an April 2014 begin-benefit-date, it notes that that's 47 months prior to FRA. So it applies a reduction of 29.5833%. In other words, it is telling you that the wife would receive 70.4166% of 50% of the husband's PIA (i.e., 35.2% of his PIA).

But that will only be the correct answer if the wife has no retirement benefit of her own. If she has a retirement benefit of her own (even if it's much smaller than the husband's benefit), she'd get a total that's somewhat more than what the calculator tells you, because the monthly reduction for early entitlement to retirement benefits is less than the monthly reduction for early entitlement to spousal benefits.

As a separate point, my math did include an error. So thanks for bringing this to my attention. Specifically, it was an error that resulted in a larger-than-correct benefit for anybody who filed early, then suspended benefits. (More specifically, I was applying DRCs to the PIA rather than to the reduced benefit amount.)

So, for anybody else who is running such scenarios (i.e., in which one person filed for benefits before FRA but is still younger than 70 such that suspension options will be tested -- Watty for instance), I'd encourage you to run the calculator again, and I apologize for the mistake. (Depending on how early the person originally filed, as well as the difference between the two people's PIAs, you may now see a different recommended strategy.)

This is particularly embarrassing, because I've made this exact error before, in a different context. (While writing the first edition of my book -- sscritic caught it in the draft.)

Also somewhat frustrating, in that the calculator had 101 different internal tests, and of course the thing I messed up is something for which I hadn't written a test. (There are now 102 tests...)

(Edited to fix a typo.)
Last edited by ObliviousInvestor on Tue Aug 14, 2018 9:26 pm, edited 1 time in total.
Mike Piper, author/blogger

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Tue Aug 14, 2018 11:07 am

Watty wrote:
Mon Aug 13, 2018 7:08 pm
On the question of when to update the numbers one thing I would keep in mind is that people will likely print off the results and show it to someone else so that the context of what has been changed will be lost. It the would be confusing to not update or blank-out the second sets of results when the top calculations are changed.
Now, if a person:
1) Runs the original calculation function, then
2) Provides alternative dates and runs the function to get the PV for those, then
3) Changes an input to the primary form (e.g., discount rate) and runs the primary calculation (i.e., top submit button),

...then in addition to the recommended strategy (and corresponding PV and table) being updated, the output for the custom dates below will be updated also.

(This is basically what we did with the bottom button yesterday. I just wasn't sure I'd be able to figure out how to do it here too. So essentially, if both forms have been filled out, then a primary input is changed, either submit button will recalculate everything, so everything is up-to-date with that new input.)

Thanks again to everybody who pointed out the ways in which the prior setup was less than desirable and who helped come up with a better/more intuitive way for this to work.
Mike Piper, author/blogger

Keith_Simons
Posts: 3
Joined: Sun Aug 12, 2018 11:31 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Keith_Simons » Tue Aug 14, 2018 7:51 pm

ObliviousInvestor wrote:
Tue Aug 14, 2018 7:59 am
Keith_Simons wrote:
Mon Aug 13, 2018 6:39 pm
here is some relevant data, without revealing personal information.
First claimant - Female - FRA March 2018, first claimed April 2014 (age 62) - PIA 920 at FRA
Second claimant - Male - FRA July 2021, entered as intending to work through end 12-2023, with a PIA at FRA of 2700, earnings per mo (flat rate) at 10K per month, not yet filing for benefits.
Other variables standard (Mortality Tables, PV interest rate)

Calculator strategy recommends:
First suspends retirement benefit till 1/2020, at age 67 and 10 months.
Your spouse files for his/her retirement benefit to begin 5/2025, at age 70 and 0 months.
You file for your spousal benefit to begin 5/2025, at age 73 and 2 months.
Yielding annual benefits for first claimant totaling 14137, of which 4560 is the spousal benefit component, the total representing 34.6% of secondary claimant's annual benefit 40768.
SSA calculator indicates that, regardless of first claimant's voluntary suspension, ratio of first claimant's total benefit to second claimant's benefit should be 35.20%.

Using a comparative claiming strategy of 1-2024 for spousal and spouse's claiming dates yields annualized benefits for first claimant that are unchanged at 14137, of which 4560 is the added spousal component. Second claimant's annualized benefit is 37440, making first claimant's total benefit 37.76% of second claimant's annual benefit.

Shouldn't the ratio be identical, regardless of the spousal claiming date? Shouldn't the ratio agree with the 35.20% calculated at SSA?
I wasn't familiar with that calculator on the SSA site, so I had to play with it a little. As it turns out, it isn't doing a very thorough analysis. All it does is assume that you get 50% of your spouse's PIA, and then applies any reduction for early entitlement. The reduction for early entitlement to spousal benefits is 25/36 of 1% for each month early, up to 36 months. For each month in excess of 36 months, the reduction is 5/12 of 1%.

So, in this case, if you enter a March 1952 DoB and an April 2014 begin-benefit-date, it notes that that's 47 months prior to FRA. So it applies a reduction of 29.5833%. In other words, it is telling you that the wife would receive 70.4166% of 50% of the husband's PIA (i.e., 35.2% of his PIA).

But that will only be the correct answer if the wife has no retirement benefit of her own. If she has a retirement benefit of her own (even if it's much smaller than the husband's benefit), she'd get a total that's somewhat more than what the calculator tells you, because the monthly reduction for early entitlement to retirement benefits is less than the monthly reduction for early entitlement to spousal benefits.

As a separate point, my math did include an error. So thanks for bringing this to my attention. Specifically, it was an error that resulted in a larger-than-correct benefit for anybody who filed early, then suspended benefits. (More specifically, I was applying DRCs to the PIA rather than to the reduced benefit amount.)

So, for anybody who else is running such scenarios (i.e., in which one person filed for benefits before FRA but is still younger than 70 such that suspension options will be tested -- Watty for instance), I'd encourage you to run the calculator again, and I apologize for the mistake. (Depending on how early the person originally filed, as well as the difference between the two people's PIAs, you may now see a different recommended strategy.)

This is particularly embarrassing, because I've made this exact error before, in a different context. (While writing the first edition of my book -- sscritic caught it in the draft.)

Also somewhat frustrating, in that the calculator had 101 different internal tests, and of course the thing I messed up is something for which I hadn't written a test. (There are now 102 tests...)
I do think you should be less hard on yourself. The key in all things such as this is the manner in which an error is addressed. Which, in your case, would be quickly, transparently and correctly.

I'm happy to have helped you, no matter how clumsy my earlier description of the problem might have been.

You should know that I have sent a message to one of the bloggers (Dana Anspach) at The Balance suggesting that they reference your calculator as a stand-in for SSAnalyze! - the loss of which inspired you to begin this project.

Ideally, they would feature the fruits of your labor atop the piece, exactly the position of prominence previously enjoyed by Bedrock's calculator.

You deserve that.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Tue Aug 14, 2018 9:21 pm

Keith_Simons wrote:
Tue Aug 14, 2018 7:51 pm
You should know that I have sent a message to one of the bloggers (Dana Anspach) at The Balance suggesting that they reference your calculator as a stand-in for SSAnalyze! - the loss of which inspired you to begin this project.
Thank you for doing that. I really appreciate it.

FWIW, I'm a big fan of Dana Anspach. In case you (or other people following this thread) haven't read it, her book Control Your Retirement Destiny is likely the best nitty-gritty-details book on retirement planning that I have read.
Mike Piper, author/blogger

User avatar
LadyGeek
Site Admin
Posts: 48102
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by LadyGeek » Wed Aug 15, 2018 8:15 pm

When selecting status as divorced*, "Your Ex-spouse's Information" retirement benefit date presents two conflicting choices.
Month in which your ex-spouse filed (or will file) for his/her retirement benefit: (end of line)
Month in which your spouse filed for his/her retirement benefit: [Month drop-down][Year drop-down]
I wasn't able to figure out which one was intended. I was randomly choosing dates to get past this error:
Please enter a later date. You cannot file for retirement benefits before the first month in which you are 62 for the entire month.
After finding something that worked, I was still left with an error message:
Your ex-spouse's filing date is necessary in order to run the calculation. If you do not know the answer, you can call the SSA to see if your ex-spouse has filed for his/her retirement benefit. Alternatively, you can guess as to your ex-spouse's plans (e.g., assume they file at age 66).
There's no place for me enter this info. I don't know what's wrong, but I think some selection drop-down boxes are missing.

I do get a Recommended Strategy and Year-by-Year Benefit Amounts, but the above description is written as an error message (further information is needed to run the calculation...) and is why I am confused and think something is wrong.

* Does not reflect my real-life situation, just testing.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
Peter Foley
Posts: 4546
Joined: Fri Nov 23, 2007 10:34 am
Location: Lake Wobegon

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Peter Foley » Wed Aug 15, 2018 8:25 pm

tc101 wrote:
Given the factors I mentioned in the last post, I wonder how useful this calculator is? It might lead to more confusion than clarity for most people, and might lead to some people taking SS earlier than would be best for them. What it calculates is just one minor part of the info to be considered when deciding to take SS.
I found it very helpful in terms of coordinating benefits between my wife and me. It had her filing a couple years sooner with good reason.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Wed Aug 15, 2018 8:26 pm

LadyGeek wrote:
Wed Aug 15, 2018 8:15 pm
When selecting status as divorced*, "Your Ex-spouse's Information" retirement benefit date presents two conflicting choices.
Month in which your ex-spouse filed (or will file) for his/her retirement benefit: (end of line)
Month in which your spouse filed for his/her retirement benefit: [Month drop-down][Year drop-down]
I wasn't able to figure out which one was intended. I was randomly choosing dates to get past this error:
Please enter a later date. You cannot file for retirement benefits before the first month in which you are 62 for the entire month.
After finding something that worked, I was still left with an error message:
Your ex-spouse's filing date is necessary in order to run the calculation. If you do not know the answer, you can call the SSA to see if your ex-spouse has filed for his/her retirement benefit. Alternatively, you can guess as to your ex-spouse's plans (e.g., assume they file at age 66).
There's no place for me enter this info. I don't know what's wrong, but I think some selection drop-down boxes are missing.

I do get a Recommended Strategy and Year-by-Year Benefit Amounts, but the above description is written as an error message (further information is needed to run the calculation...) and is why I am confused and think something is wrong.

* Does not reflect my real-life situation, just testing.
Ah...interesting. Apparently this is what happens if you first choose "married," select "yes" for your spouse having filed, then switch marital status to "divorced."

Thanks for the heads-up. I'll see if I can get this fixed tomorrow.

(The solution for now, for anybody else who gets stuck like this -- until I fix the code so this problem doesn't happen -- is to switch back to "married" and choose "no" for whether your spouse has filed or not.)
Mike Piper, author/blogger

User avatar
LadyGeek
Site Admin
Posts: 48102
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by LadyGeek » Wed Aug 15, 2018 8:38 pm

^^^ Not from switching status, but from a clean start. From a new private browser window (clean browser cache):

First choose Marital status "Divorced".

Change nothing else except:

Month in which your ex-spouse filed (or will file) for his/her retirement benefit: 5 - 1991 I only see this single line entry at this point.

Error: Please enter a later date. You cannot file for retirement benefits before the first month in which you are 62 for the entire month.

Change the "Your Ex-spouse's Information" Date of birth to 4 - 15 - 1947 (just change the year).

Two lines now appear:
Month in which your ex-spouse filed (or will file) for his/her retirement benefit:
Month in which your spouse filed for his/her retirement benefit:[month][year]

Change the 2nd line year to:
Month in which your spouse filed for his/her retirement benefit: [ 05 ][ 2012 ]

The calculation completes with the error message:
Your ex-spouse's filing date is necessary in order to run the calculation. If you do not know the answer, you can call the SSA to see if your ex-spouse has filed for his/her retirement benefit. Alternatively, you can guess as to your ex-spouse's plans (e.g., assume they file at age 66).

The Recommended Strategy report is displayed.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Wed Aug 15, 2018 8:51 pm

There's apparently two different things causing the same problem. (Or rather, two different ways to trigger the same bug in the code.)

Essentially the bug gets triggered whenever marital status is divorced and the filing date of "Person B" is considered to be fixed (through an input other than the normal input for ex-spouse filing date). This happens if they're over 70 -- because anybody over 70 is considered to have a fixed filing date, since they have presumably filed by now -- or if the "yes they have filed" input was selected when "married" was selected as marital status.

Thanks for the detailed description of what you're seeing and how you got there. Much appreciated! (I don't think it'll be hard to fix. Should be able to knock it out tomorrow.)

Edited to add: just for reference, that message isn't an error message. It's just a permanent message when "divorced" is selected, so that the user knows that the input can't just be skipped (as one might be inclined to try if they don't have such information available), and to give some ideas about how to get that information. I've tried to follow an "error messages with red background, informational messages with blue background" convention.
Mike Piper, author/blogger

User avatar
LadyGeek
Site Admin
Posts: 48102
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by LadyGeek » Wed Aug 15, 2018 10:02 pm

ObliviousInvestor wrote:
Wed Aug 15, 2018 8:51 pm
Edited to add: just for reference, that message isn't an error message. It's just a permanent message when "divorced" is selected, so that the user knows that the input can't just be skipped (as one might be inclined to try if they don't have such information available), and to give some ideas about how to get that information. I've tried to follow an "error messages with red background, informational messages with blue background" convention.
Now I see what's wrong. This is an information message explaining why the input - to be entered below - is needed.

Place that text before the entry so I can see that it is an instruction on what to do next vs. something I've already done and is after the fact.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Thu Aug 16, 2018 3:28 pm

LadyGeek, I believe I have fixed the issues you raised.

Relatedly, I also made it so that the message about the ex-spouse's filing date now is an error message. (That is, by default there's no message there now. But the message now shows up as an error message if they leave the field blank.)
Mike Piper, author/blogger

User avatar
One Ping
Posts: 488
Joined: Thu Sep 24, 2015 4:53 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by One Ping » Thu Aug 16, 2018 5:53 pm

I was showing a friend the tool and something happened with his case that I didn't expect.

Husband:
Birthday: 2/15/1954 (Note: He is not eligible to file a restricted application)
PIA: 2500
Wife:
Birthday: 8/15/1953 (Note: She is eligible to file a restricted application)
PIA: 1000

The tool says the 'best' strategy is:
Wife files for her benefit on 8/2018, age 65yr 0mo
Husband files for his benefit on 2/2024, age 70yr 0 mo
Wife files for her spousal benefit on 2/2024 (i.e., now), age 70yr 6mo.

So far, so good ...

Then we wanted to run this alternative case :
Husband claims his benefit on 8/2018
Wife claims her benefit on 8/2023
Wife claims her spousal benefit on 8/2018 (i.e., restricted spousal application)

Instead of getting a result, we get a message next to the spousal claim date box that says:
"You can't file a restricted application (i.e., application for spousal-only) prior to your FRA."

I thought you could file for spousal benefits at anytime, if your spouse had filed for their benefit. Is the problem here that her application is a 'restricted' application because she is not filing for her 'own' benefit at that time and you really can't do that before FRA?

One Ping
"Re-verify our range to target ... one ping only."

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Thu Aug 16, 2018 6:05 pm

One Ping wrote:
Thu Aug 16, 2018 5:53 pm
I thought you could file for spousal benefits at anytime, if your spouse had filed for their benefit.
Yes, as long as you are 62.
One Ping wrote:
Thu Aug 16, 2018 5:53 pm
Is the problem here that her application is a 'restricted' application because she is not filing for her 'own' benefit at that time and you really can't do that before FRA?
Yes, that's exactly the problem. An application for spousal-only is a restricted application, and that cannot be done prior to FRA.

See D(1)(b) here:
https://secure.ssa.gov/poms.nsf/lnx/0200204020
Mike Piper, author/blogger

User avatar
One Ping
Posts: 488
Joined: Thu Sep 24, 2015 4:53 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by One Ping » Thu Aug 16, 2018 6:08 pm

Thanks, Mike. :beer

So, since the recent change to the law, you can't really file for your spousal benefit before FRA unless you take your own benefit at the same time (deemed filing?), right?
"Re-verify our range to target ... one ping only."

vtMaps
Posts: 400
Joined: Tue Oct 14, 2008 12:05 pm
Location: central Vermont

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by vtMaps » Thu Aug 16, 2018 7:58 pm

One Ping wrote:
Thu Aug 16, 2018 6:08 pm
So, since the recent change to the law, you can't really file for your spousal benefit before FRA unless you take your own benefit at the same time (deemed filing?), right?
Even before the law changed, I don't think you could file a restricted application before your FRA.

--vtMaps
The optimist proclaims that we live in the best of all possible worlds; and the pessimist fears this is true. --James Branch Cabell

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Thu Aug 16, 2018 8:14 pm

vtMaps wrote:
Thu Aug 16, 2018 7:58 pm
One Ping wrote:
Thu Aug 16, 2018 6:08 pm
So, since the recent change to the law, you can't really file for your spousal benefit before FRA unless you take your own benefit at the same time (deemed filing?), right?
Even before the law changed, I don't think you could file a restricted application before your FRA.

--vtMaps
vtMaps is correct.

What the law changed is that for those of us born after 1/1/1954, a restricted application is not possible at any age.

(A restricted application for spousal benefits, that is. Restricted applications for survivor benefits are still available beginning as early as age 60, regardless of DoB.)
Mike Piper, author/blogger

User avatar
LadyGeek
Site Admin
Posts: 48102
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by LadyGeek » Thu Aug 16, 2018 8:39 pm

ObliviousInvestor wrote:
Thu Aug 16, 2018 3:28 pm
LadyGeek, I believe I have fixed the issues you raised.

Relatedly, I also made it so that the message about the ex-spouse's filing date now is an error message. (That is, by default there's no message there now. But the message now shows up as an error message if they leave the field blank.)
I repeated my previous test case from this post (9 posts up). When the ex-spouse fields are left blank, I do get the error message - which makes total sense now.

You need to add the Year to the ex-spouse instructions. There are entry boxes, so it should be clear on the meaning of each. From:
Month in which your ex-spouse filed (or will file) for his/her retirement benefit:
To:
Month and Year in which your ex-spouse filed (or will file) for his/her retirement benefit:
I see you are testing for either box to be blank, which is good.

====================
Alternative suggestion: Provide the date format to the right of each input description (left of the entry boxes). I like this better - it's consistent and clear. It also eliminates guess work.

On a cellphone, I think this is necessary as drop-down lists are much harder to work with than on a desktop.

In the "Your Information" and wherever a date is needed, add the format to the right the text input. So,

Code: Select all

Date of birth (mm-dd-yyyy): [drop-down] [drop-down] [drop-down]
In "Month in which your ex-spouse filed (or will file) for his/her retirement benefit:", change to:

Code: Select all

Date in which your ex-spouse filed (or will file) for his/her retirement benefit (mm-yyyy): [drop-down] [drop-down]
I highly recommend you test this with a mobile device, as the drop-down boxes will drop to the next line (responsive layout). Without guidance, it can get confusing and difficult to work with.

Tip: In Chrome --> right click --> Inspect --> Simulate Mobile Devices with Device Mode
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Fri Aug 17, 2018 7:53 am

LadyGeek wrote:
Thu Aug 16, 2018 8:39 pm
From:
Month in which your ex-spouse filed (or will file) for his/her retirement benefit:
To:
Month and Year in which your ex-spouse filed (or will file) for his/her retirement benefit:
Good suggestion. Did this this morning.

And thank you for the input regarding other "user experience" considerations for the form. It's definitely something I'm looking into. Right now I'm collecting feedback (here and from people emailing), and doing some reading on "best practices" for forms. I also have a family member who is a developer who does a lot of work with forms, so I'll probably pick her brain too.
Mike Piper, author/blogger

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Thu Sep 06, 2018 10:16 am

Rolled out a few updates today:

1) The calculator now allows a person to note if they are on disability (and expecting to stay on disability until FRA), and it adjusts calculations and output accordingly.
2) There's now a link at the bottom of the output to print the page (which can also be used to save as pdf). This option is just a normal print page option, excluding some elements that I figured people wouldn't want in the saved output (e.g., the page footer, the note explaining what a PIA is).
3) Made a change that cut calculation time in Chrome/Firefox by ~30% and Safari by ~80%.
Mike Piper, author/blogger

47Percent
Posts: 337
Joined: Tue Oct 29, 2013 9:59 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by 47Percent » Fri Sep 07, 2018 6:15 pm

ObliviousInvestor wrote:
Wed Jun 06, 2018 7:20 am

One thing that would help me to make this calculator as useful as possible: if you have input about how you want these features to be implemented, that would be helpful.

For instance, with regard to any of the above features, what would you like the user interface to look like?
Thanks for your really helpful book and this wonderful work.
I can't believe such a tool is not available at the ssa website. It ought to be!

A few suggestions:'

1) Please include a way to include an actual expected lifetime. For example if I expect to live up to 90 due to family history etc. it would be nice to be able to just enter that number instead of going by life expectancy table.
My bad.. As pointed out later, I missed the fact that this is already available!!! (Thanks @rkhusky)

2) Please keep the current method of entering actual birth dates. i.e. do not switch to entering as age.

3) I am eagerly waiting for your Child addition. I still have 2 years. So hopefully it will be there before that!!
Last edited by 47Percent on Fri Sep 07, 2018 10:17 pm, edited 4 times in total.

rkhusky
Posts: 5554
Joined: Thu Aug 18, 2011 8:09 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by rkhusky » Fri Sep 07, 2018 8:39 pm

47Percent wrote:
Fri Sep 07, 2018 6:15 pm

1) Please include a way to include an actual expected lifetime. For example if I expect to live up to 90 due to family history etc. it would be nice to be able to just enter that number instead of going by life expectancy table.
This is already included in advanced mode (Advanced options checkbox at the top) and the last option in the Mortality Table drop down menu.

dconrath
Posts: 7
Joined: Fri Jun 08, 2018 10:24 am

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by dconrath » Fri Sep 07, 2018 9:57 pm

See the following arguing that taking into consideration taxation issues, it sometimes makes more sense to take Social Security at an earlier age:
https://www.kiplinger.com/article/retir ... at-62.html

I'm not sure there is really enough details in the article to determine if the conclusion is correct or not.

If the conclusion does have some degree of validity, is there anyway to take into consideration these taxation issues in the evaluations the "Open Social Security" tool calculates?

I can't say thank you enough for the effort involved in producing this tool and supporting it's use and answering questions and comments!

Thank you :-)

User avatar
FiveK
Posts: 5613
Joined: Sun Mar 16, 2014 2:43 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by FiveK » Fri Sep 07, 2018 10:27 pm

dconrath wrote:
Fri Sep 07, 2018 9:57 pm
I'm not sure there is really enough details in the article to determine if the conclusion is correct or not.
Doesn't look correct at first glance. The article claims
If this family were to delay Social Security payments from age 62 to age 66, it’s estimated they would be deferring nearly $146,000 in Social Security payments over that period. They would be forced to pull money from their retirement accounts to live on, paying a projected $51,372 in federal taxes from age 62 to 66.

Instead, if they elected to take their Social Security at age 62 at a 10% bracket, the tax savings on that $146,000 would be substantial. By drawing early and minimizing taxes on retirement and brokerage funds, the family would owe $9,768 in taxes during the same period. That’s a savings of $41,604 in taxes at the beginning of their retirement career.
$146K/4 years = $36.5K/yr

$51,372/4 years = $12,843 federal tax/yr (presumably federal income tax)

It takes ~$119,291 in tIRA withdrawals alone for a 63 year old couple to pay $12,843.

Replacing $36.5K of tIRA withdrawals with $36.5K of SS benefits (so $82,791 tIRA) incurs federal income tax of $11,639/yr. That's a lot more than $9,768 over four years - more than would plausibly change if a sharpened pencil included changes in standard deduction at age 65, etc.

Unless I misread something - always possible - the example doesn't bear scrutiny.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Fri Sep 07, 2018 10:41 pm

47Percent wrote:
Fri Sep 07, 2018 6:15 pm
3) I am eagerly waiting for your Child addition. I still have 2 years. So hopefully it will be there before that!!
I've already been working on it for a while, but it looks like it's going to be a big rework. The coding for disability benefits ended up taking a couple months, and I wouldn't be surprised if this is similar. (Though the honest truth is, it could be much shorter or much longer. I'm just not very good at estimating these things yet.)

On the topic of using mortality tables as opposed to a fixed death-age assumption, I get that it's helpful to have the fixed-death-age option if you have a specific health condition or if you just want to run "what if" analyses. But I would at least urge any married people to also do the analysis with various mortality tables. Fixed death-age assumptions have the flaw of understating the value of the high-PIA spouse delaying and overstating the value of the low-PIA spouse delaying.
dconrath wrote:
Fri Sep 07, 2018 9:57 pm
See the following arguing that taking into consideration taxation issues, it sometimes makes more sense to take Social Security at an earlier age:
https://www.kiplinger.com/article/retir ... at-62.html

I'm not sure there is really enough details in the article to determine if the conclusion is correct or not.

If the conclusion does have some degree of validity, is there anyway to take into consideration these taxation issues in the evaluations the "Open Social Security" tool calculates?
I only quickly read the article just now. Did not check its math or deeply think about its analysis. (Edit: I see FiveK has beat me to it.)

But what I can say is that tax planning is always case-by-case, including the tax-planning aspect of Social Security, because almost anything that appears on a person's 1040 could be an important factor in the analysis.

In most cases I have looked at though (and when I say most, I just mean "more than 50%," not "almost all"), tax considerations have pointed in favor of delaying. The overall reason being that spending down tax-deferred accounts in order to delay Social Security has the effect of reducing the percentage of your lifetime income that is made up of distributions from tax-deferred accounts (which are usually fully taxable) and increasing the percentage of your lifetime income that is made up of Social Security benefits (which are at most 85% taxable, sometimes much less).

And, unfortunately, I just don't see any way of including this in the Open Social Security calculator. A good analysis requires something along the lines of TurboTax or other sophisticated tax software, the creation of which is well beyond my capabilities.
Last edited by ObliviousInvestor on Sat Sep 08, 2018 9:37 am, edited 2 times in total.
Mike Piper, author/blogger

dconrath
Posts: 7
Joined: Fri Jun 08, 2018 10:24 am

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by dconrath » Fri Sep 07, 2018 11:01 pm

FiveK wrote:
Fri Sep 07, 2018 10:27 pm

Doesn't look correct at first glance. The article claims
If this family were to delay Social Security payments from age 62 to age 66, it’s estimated they would be deferring nearly $146,000 in Social Security payments over that period. They would be forced to pull money from their retirement accounts to live on, paying a projected $51,372 in federal taxes from age 62 to 66.

Instead, if they elected to take their Social Security at age 62 at a 10% bracket, the tax savings on that $146,000 would be substantial. By drawing early and minimizing taxes on retirement and brokerage funds, the family would owe $9,768 in taxes during the same period. That’s a savings of $41,604 in taxes at the beginning of their retirement career.
$146K/4 years = $36.5K/yr

$51,372/4 years = $12,843 federal tax/yr (presumably federal income tax)

It takes ~$119,291 in tIRA withdrawals alone for a 63 year old couple to pay $12,843.

Replacing $36.5K of tIRA withdrawals with $36.5K of SS benefits (so $82,791 tIRA) incurs federal income tax of $11,639/yr. That's a lot more than $9,768 over four years - more than would plausibly change if a sharpened pencil included changes in standard deduction at age 65, etc.

Unless I misread something - always possible - the example doesn't bear scrutiny.
Thanks FiveK, for your quick look and feedback which supports what I had been thinking. The numbers they mentioned seemed to be related to nonequivalent substitutions and with what was offered in the details just didn't seem to make sense to me.
ObliviousInvestor wrote:
Fri Sep 07, 2018 10:41 pm

I only quickly read the article just now. Did not check its math or deeply think about its analysis. (Edit: I see FiveK has beat me to it.)

But what I can say is that tax planning is always case-by-case, including the tax-planning aspect of Social Security, because almost anything that appears on a person's 1040 could be an important factor in the analysis.

In most cases I have looked at though (and when I say most, I just mean "more than 50%," not "almost all"), tax considerations have pointed in favor of delaying. The overall reason being that spending down tax-deferred accounts in order to delay Social Security has the effect of reducing the percentage of your lifetime income that is made up of distributions from tax-deferred accounts (which are usually fully taxable) and increasing the percentage of your lifetime income that is made up of Social Security benefits (which are at most 85% taxable, sometimes much less).

And, unfortunately, I just don't see any way of including this in the Open Social Security calculator. A good analysis requires something along the lines of TurboTax or other sophisticated tax software, the creation of which is well beyond my capabilities.
I appreciate your generalized feedback that tax considerations have often pointed in favor of delaying. Your explanation makes sense to me...

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Sat Sep 08, 2018 9:36 am

To elaborate somewhat now that I have more time:

The mechanism at work that usually makes delaying tax-advantageous is that trading a dollar of (traditional) IRA income for a dollar of Social Security income is usually beneficial, because Social Security is at most 85% taxable.

A second part that sometimes is relevant is that the "combined income" figure that determines how much of your Social Security is taxable only includes 50% of your Social Security benefits, whereas it includes 100% of all your other sources of taxable income (plus tax-exempt interest). So trading a dollar of IRA income for a dollar of Social Security income can even make the tax-advantaged Social Security income even more tax-advantaged.

A third part that can sometimes be relevant is that since now you (maybe) have a lower portion of your Social Security included in taxable income, you might even be paying less tax on your IRA income (i.e., you've made the IRA income more tax-advantaged as well).

But #2 doesn't apply at all for some people, as they might have no way at all to get it below 85%-taxable.

And #3 is less likely to apply if #2 doesn't apply.

And there can be any number of other things to consider, which could push the decision in either direction. A few common such things would be tax planning for ACA subsidies, Medicare premiums, or an expected inheritance of a sizable IRA -- but really it could be anything.
Mike Piper, author/blogger

Dottie57
Posts: 4507
Joined: Thu May 19, 2016 5:43 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Dottie57 » Mon Sep 10, 2018 3:45 pm

Hi,

There appears to be a problem which I have not seen before


2026 = 28k
2027 = 48k

Big jump for 1 year!

I put in a PIA amount of 3300, birth date in 1957 and gebder of female.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Sep 10, 2018 3:47 pm

Dottie57 wrote:
Mon Sep 10, 2018 3:45 pm
Hi,

There appears to be a problem which I have not seen before


2026 = 28k
2027 = 48k

Big jump for 1 year!

I put in a PIA amount of 3300, birth date in 1957 and gender of female.
What is the problem? It looks to me like it's probably recommending a claiming date in June 2026. So you get 7 months of benefits in 2026, and 12 months of benefits in 2027.
Mike Piper, author/blogger

Dottie57
Posts: 4507
Joined: Thu May 19, 2016 5:43 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Dottie57 » Mon Sep 10, 2018 3:54 pm

ObliviousInvestor wrote:
Mon Sep 10, 2018 3:47 pm
Dottie57 wrote:
Mon Sep 10, 2018 3:45 pm
Hi,

There appears to be a problem which I have not seen before


2026 = 28k
2027 = 48k

Big jump for 1 year!

I put in a PIA amount of 3300, birth date in 1957 and gender of female.
What is the problem? It looks to me like it's probably recommending a claiming date in June 2026. So you get 7 months of benefits in 2026, and 12 months of benefits in 2027.
Ok. Didn’t realize you showed half years.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Mon Sep 10, 2018 3:59 pm

Dottie57 wrote:
Mon Sep 10, 2018 3:54 pm
Didn’t realize you showed half years.
Thank you for taking the time to point out the potential source of confusion though. If you misunderstood it, you're surely not the only one. I'll see if I can come up with something to make the output more clear.
Mike Piper, author/blogger

You Know What I Mean
Posts: 11
Joined: Tue Mar 06, 2018 9:27 am

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by You Know What I Mean » Fri Sep 14, 2018 7:38 pm

Mike,

Thanks for the great calculator and for recently adding the helpful Disability option, which is very timely!

If a person is on SS disability, which converts to a retirement benefit of the same amount at FRA, can that person suspend the retirement benefit at FRA and apply for survivor benefit at that time, allowing the retirement benefit to get delayed retirement credits (DRC’s) till age 70?

Here are the specifics on the situation. I am helping a woman who was born in March 1955. I had previously planned that as soon as she retires – probably at age 65 – she would apply for survivor benefits. That would be about $1113/month at age 65 (or about $1163 at age 66, which is FRA for survivor benefits). Unless she is in poor health, I was planning that she would defer her own retirement benefit till age 70.

However, as of January 2018 she is on SS disability of $1946/month. She may be on it till her FRA of 66 years, 2 months. Your calculator recommends suspending her retirement benefit then and allowing the DRC’s to accrue until age 69 years, 1 month. Makes sense. It would be even better if she could get the survivor benefits for the 2 years, 11 months of suspension. Is that possible? If it is, I imagine it might even be better to unsuspend at 70 rather than 69 years, 1 month. Would that calculation be another advanced option for your calculator?

Thanks again for the calculator and for all the advice you have provided. I read and enjoyed your Social Security Made Simple book when it first came out – very helpful.

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Fri Sep 14, 2018 9:01 pm

You Know What I Mean wrote:
Fri Sep 14, 2018 7:38 pm
Mike,

Thanks for the great calculator and for recently adding the helpful Disability option, which is very timely!

If a person is on SS disability, which converts to a retirement benefit of the same amount at FRA, can that person suspend the retirement benefit at FRA and apply for survivor benefit at that time, allowing the retirement benefit to get delayed retirement credits (DRC’s) till age 70?

Here are the specifics on the situation. I am helping a woman who was born in March 1955. I had previously planned that as soon as she retires – probably at age 65 – she would apply for survivor benefits. That would be about $1113/month at age 65 (or about $1163 at age 66, which is FRA for survivor benefits). Unless she is in poor health, I was planning that she would defer her own retirement benefit till age 70.

However, as of January 2018 she is on SS disability of $1946/month. She may be on it till her FRA of 66 years, 2 months. Your calculator recommends suspending her retirement benefit then and allowing the DRC’s to accrue until age 69 years, 1 month. Makes sense. It would be even better if she could get the survivor benefits for the 2 years, 11 months of suspension. Is that possible? If it is, I imagine it might even be better to unsuspend at 70 rather than 69 years, 1 month. Would that calculation be another advanced option for your calculator?
Unfortunately that is not possible. In such a scenario she would get no survivor benefit. (Reason being, she would still be considered "entitled" to her retirement benefit even if it is suspended. So her survivor benefit would be reduced by the amount of her retirement benefit -- all the way to zero.)

She does have the option of withdrawing her disability application (and paying back benefits received so far) and filing for just survivor benefits -- and collecting those from now until 70.

If she stays on disability she will have collected a total of 40 months of disability before FRA (40 x $1946 = $77,840).

If she withdraws her application for disability and applies immediately for survivor benefits, she'd get 78 months of such until age 70. (78 x $1113 = $86,814)

So in theory there's a little bit of room for improvement here. But it would involve paying back a lump sum (plus doing the withdrawal of application paperwork, as well as a new application). And every month of delay would take $1113 out of that room for improvement.

(Also, I only counted once, so it's possible I messed up either of these month counts.)
Mike Piper, author/blogger

You Know What I Mean
Posts: 11
Joined: Tue Mar 06, 2018 9:27 am

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by You Know What I Mean » Sat Sep 15, 2018 6:11 am

ObliviousInvestor wrote:
Fri Sep 14, 2018 9:01 pm
You Know What I Mean wrote:
Fri Sep 14, 2018 7:38 pm
Mike,

Thanks for the great calculator and for recently adding the helpful Disability option, which is very timely!

If a person is on SS disability, which converts to a retirement benefit of the same amount at FRA, can that person suspend the retirement benefit at FRA and apply for survivor benefit at that time, allowing the retirement benefit to get delayed retirement credits (DRC’s) till age 70?

Here are the specifics on the situation. I am helping a woman who was born in March 1955. I had previously planned that as soon as she retires – probably at age 65 – she would apply for survivor benefits. That would be about $1113/month at age 65 (or about $1163 at age 66, which is FRA for survivor benefits). Unless she is in poor health, I was planning that she would defer her own retirement benefit till age 70.

However, as of January 2018 she is on SS disability of $1946/month. She may be on it till her FRA of 66 years, 2 months. Your calculator recommends suspending her retirement benefit then and allowing the DRC’s to accrue until age 69 years, 1 month. Makes sense. It would be even better if she could get the survivor benefits for the 2 years, 11 months of suspension. Is that possible? If it is, I imagine it might even be better to unsuspend at 70 rather than 69 years, 1 month. Would that calculation be another advanced option for your calculator?
Unfortunately that is not possible. In such a scenario she would get no survivor benefit. (Reason being, she would still be considered "entitled" to her retirement benefit even if it is suspended. So her survivor benefit would be reduced by the amount of her retirement benefit -- all the way to zero.)

She does have the option of withdrawing her disability application (and paying back benefits received so far) and filing for just survivor benefits -- and collecting those from now until 70.

If she stays on disability she will have collected a total of 40 months of disability before FRA (40 x $1946 = $77,840).

If she withdraws her application for disability and applies immediately for survivor benefits, she'd get 78 months of such until age 70. (78 x $1113 = $86,814)

So in theory there's a little bit of room for improvement here. But it would involve paying back a lump sum (plus doing the withdrawal of application paperwork, as well as a new application). And every month of delay would take $1113 out of that room for improvement.

(Also, I only counted once, so it's possible I messed up either of these month counts.)
Mike,

Thanks so much for the quick, detailed reply!

By the way, as a practical matter she can't really withdraw her Social Security disability application. She is also on long-term disability (LTD) through an insurance policy from work that requires her to apply for Social Security disability. In fact, she has to re-pay the insurance company for all of the SSDI lump sum ($13,622) that she just received for the first seven months of this year. Going forward, the insurance company's LTD payments to her are reduced by her $1,946 SSDI benefit.

User avatar
Watty
Posts: 14135
Joined: Wed Oct 10, 2007 3:55 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Watty » Fri Sep 28, 2018 9:04 pm

Hi,

I have a question about the way it does the calculation when it compares "Test an Alternative Strategy" to the recommended strategy figure out the difference. For example mine says;
The present value of the strategy you selected is $624,324, as compared to a present value of $647,761 from the recommended strategy, a difference of $23,437.
When I start at my full retirement age compared to the recommended strategy which would have me start at 70

The situation that I am looking at is that if I start at my FRA (66 years 4 months) my yearly benefit will be about $30,000 so I will not need to withdraw $30,000 a year from retirement savings. Between my FRA and the age of 70 that would be about $110,000 that I would not have taken out of my retirement account.

My question is if the money I would not withdraw from my funds by starting early is included in the present value calculation of the alternate strategy.

The actual math is much much more complex but if this was not somehow included then it would seem that starting at my FRA and having an extra $110K in retirement savings when I turn 70 would be better than starting at 70 and having a higher present value of $23K.


The way I entered it was like this if you want to play with the numbers;

Me: married,male,birth 9/15/56, pia 2500
Spouse Female, birth 6/15/51, pia 1000 started 7/2013

Alternative strategy; me - claim 1/2023
Do not suspend.
Spouse claim 1/2023

User avatar
ObliviousInvestor
Posts: 3387
Joined: Tue Mar 17, 2009 9:32 am
Contact:

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by ObliviousInvestor » Fri Sep 28, 2018 9:30 pm

Watty wrote:
Fri Sep 28, 2018 9:04 pm
Hi,

I have a question about the way it does the calculation when it compares "Test an Alternative Strategy" to the recommended strategy figure out the difference. For example mine says;
The present value of the strategy you selected is $624,324, as compared to a present value of $647,761 from the recommended strategy, a difference of $23,437.
When I start at my full retirement age compared to the recommended strategy which would have me start at 70

The situation that I am looking at is that if I start at my FRA (66 years 4 months) my yearly benefit will be about $30,000 so I will not need to withdraw $30,000 a year from retirement savings. Between my FRA and the age of 70 that would be about $110,000 that I would not have taken out of my retirement account.

My question is if the money I would not withdraw from my funds by starting early is included in the present value calculation of the alternate strategy.
The calculation is of present value of expected benefits received over your and your spouse's lifetimes. So, yes, it is accounting for the "opportunity cost" you incur when you delay benefits. The discount rate reflects the rate of return that you are giving up on the money you are spending down (i.e., money that would have stayed invested, in a file-early scenario).
Mike Piper, author/blogger

User avatar
Watty
Posts: 14135
Joined: Wed Oct 10, 2007 3:55 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by Watty » Sat Sep 29, 2018 10:32 am

ObliviousInvestor wrote:
Fri Sep 28, 2018 9:30 pm
The calculation is of present value of expected benefits received over your and your spouse's lifetimes. So, yes, it is accounting for the "opportunity cost" you incur when you delay benefits. The discount rate reflects the rate of return that you are giving up on the money you are spending down (i.e., money that would have stayed invested, in a file-early scenario).

I'm still trying to understand how this works and I must be missing something.

In the example I gave it said, "The present value of the strategy you selected is $624,324.." which I am assuming is the sum of all the individual yearly present values adjusted for the probably that you will be alive in that year. I reread your web page explaining the present values and I think I understand that.

https://obliviousinvestor.com/what-does ... alue-mean/

On that same page though it says;
So What Does Net Present Value Mean?
The net present value of an investment/strategy is the sum of the present values of all of the cash flows received, minus the sum of the present values of all of the cash outflows. In the case of Social Security claiming strategies, however, there are no relevant cash outflows. That is, the cash outflows are the Social Security payroll taxes you pay over the course of your career in order to qualify for a benefit, but those taxes are the same regardless of which claiming strategy you use, so we do not need to include them in an analysis that compares claiming strategies.


I think where I am getting confused is that if I delay starting Social Security after my FRA then I would have to fund $30,000 a year from my retirement savings until I start Social Security at the recommended age of 70. If you are wanted to look at the "net present value", which seems to be more relevant, instead of just the "present value" then wouldn't that $30K a year be need to be considered as a "relevant cash outflow" like you described in your description?

User avatar
FiveK
Posts: 5613
Joined: Sun Mar 16, 2014 2:43 pm

Re: "Open Social Security" calculator: feature requests, bug reports, etc

Post by FiveK » Sat Sep 29, 2018 10:49 am

Watty wrote:
Sat Sep 29, 2018 10:32 am
I think where I am getting confused is that if I delay starting Social Security after my FRA then I would have to fund $30,000 a year from my retirement savings until I start Social Security at the recommended age of 70. If you are wanted to look at the "net present value", which seems to be more relevant, instead of just the "present value" then wouldn't that $30K a year be need to be considered as a "relevant cash outflow" like you described in your description?
It depends how wide one wants to draw the boundaries.

Yes, if one has $0 savings and no other income, taking SS immediately is the logical choice because having money for food is good.

If one has a large amount of traditional savings, however, deferring SS until age 70 while converting/withdrawing those savings at lower tax rates may be the best choice.

Both of those situations are outside the scope of the "highest present value of SS benefits" because that looks at the SS benefit stream in isolation. So, how wide to draw the boundaries...?

Post Reply