Taxable Social Security Benefits... Pay Now or Pay Later? | Ed Slott and Company, LLC

Taxable Social Security Benefits... Pay Now or Pay Later?

I think my parents have a "high-quality problem" and I'm hoping this forum can help me generate a few more ideas for them. Mom is 68 and draws a state government pension and Social Security benefits. Dad is 74 and draws a state government pension and Social Security benefits too. Both are on Medicare. Both are able to live modestly off their pensions and Social Security and thus have no real need to drill into their savings account. Both also have rollover IRAs and Roth IRAs. Because of dad's age, he's required to take a minimum amount yearly from his rollover IRA. They have minimal other income (like some interest off of their savings account). Both file taxes jointly.

Because of all this, I see no immediate opportunities to trim income and thus I see no immediate opportunities to get them into the 10% marginal tax bracket (currently in the 15% bracket I believe). And when it comes to how much of their Social Security is taxable, they seem to make just enough "provisional income" over $44000 that some is exposed at that 85% level. So this leaves me wondering how to handle their rollover IRAs. They don't really need those funds now or even in the near-term, but dad is forced to take RMDs (this is partly why their income ticks over $44k in provisional income), and mom will join him in a few years.

So a conversion to their Roths makes some sense. Of course we want to be mindful of avoiding the next higher marginal tax bracket, and to avoid higher Medicare premiums next year. But, for example, converting $10k from their rollover IRAs to their Roth IRAs doesn't just mean $10k extra in income to show on returns it seems. It means $18.5k extra, because another $8500 of their Social Security benefits would be exposed to taxes in this example.

That seems like an ouchie... but is this just a case of "pay it now or pay it later?" If they have no immediate options of lowering income and potentially higher income once mom's RMDs kick-in, all of it will be provisional income over that $44k bound, and thus subject to the same math now or later. Except if later in the form of a RMD, they won't have the option of getting it into the Roth for those yummy benefits, right? Am I missing anything? Any other options/ideas?

Any suggestions, from anyone?

All the income appears to be stabilized except for Mom's RMDs. Seems like the pensions must be modest or they would already be out of the 15% bracket and 85% of SS would be included in AGI. And the only flexibility they have remaining is Roth conversions. You could use Taxcaster (on line basic tax program) to determine by increment the cost of converting various increments of her TIRA, but there is only one or two tax years left before her RMDs kick in. You may find that their marginal rate for her conversions changes and will may actually drop once all 85% of the SS is included. How this works is dependent on dollar amounts of the RMDs, pensions and SS so it is near impossible to speculate. When figuring the cost of her conversions you might use increments of every 3,000 to see how much the tax bill rises. Divide the tax increase by 3,000 to get the rate for that increment, do the same for 6k, 9k etc. Convert any additional IRMAA surcharge into a tax rate etc.  Once you have these costs of various conversion amounts, the final step is determine if you can how much the future RMDs will be reduced due to the conversions and at what tax bracket. Conversion opportunities could still exist after her RMDs kick in, but those fall into higher income years because either parent can only convert additional amounts AFTER their annual RMDs have been completed. This tax modeling is pretty complex but at least the number of variables in their case is limited.

Hi, and thanks for the thoughtful response Alan, as always!  I toyed with that Taxcaster tool, and ran some numbers.  I think I see what you mean about the marginal rate on the conversions.  They did eventually come down (i.e. $20k conversion was at a marginal rate of 26.52%, and $25k conversion was at a marginal rate of 25.35%... but $5k conversion was at a marginal rate of 22.82%).  I had to Google what "IRMAA" meant.  But I knew about potentially increased Medicare premiums from other research.  But the zone where they kick in, around $170k jointly, is not a concern... they'll stay well under that I'm sure.  I haven't tried looking at future years and reduced RMDs yet... and here's why: while playing on Taxcaster, I noticed (which I probably should have known before), mom and dad are actually, probably, going to be in the 10% marginal tax bracket for 2015.  My previous analysis overlooked the deductions and exemptions that would reduce their taxable income.  So it seems that mom and dad can only convert another $2600 roughly before they do indeed move from the 10% bracket to the 15% bracket.  So having said that... does it make sense to force them into the higher bracket now, or just do what we can in the current bracket and then wait until later, once mom's RMDs kick-in and they're really into the 15% bracket for good, before doing more conversion?

Any other thoughts, from anyone?  I thought an "unofficial" rule of such things was to never bump yourself into the next higher marginal bracket.  I plan to double-check with my parents soon, but if they're in the 10% marginal bracket now after all, jumping into the 15% marginal bracket prematurely seems like a bad call.  But I still haven't gotten a handle of all the future RMD math that Alan suggests.  Seems complex enough to not rush it... but I'm going to run out of time in 2015 fast.  If staying put in the 10% bracket seems like the best option, regardless of the future situation, then I can allow myself more time to study Alan's suggest.  So again, any other thoughts?

In addition to those last questions about trying to stick to the 10% bracket and the merit of pushing into the 15% bracket in this case, I had another question.  I think I know how to estimate future RMDs for my parents.  For better or for worse, their traditional IRAs are in cash.  So I can compute what their balances will be on 31-Dec of this year pretty readily under multiple conversion scenarios.  And then I can run the math again for next year.  And then I can determine what mom's first RMD would be.  But I'm not following what this helps me understand?  Mom's traditional IRA balance now is what dad's was, roughly, when he started RMDs.  And I think his first RMD was in the $3k to $5k range.  If I assume mom's first RMD is in that range, then they'd almost certainly tick into the 15% marginal bracket then (they're only about $2500 from that bracket now).  I can see how a conversion now could potentially reduce the RMD in the future... but it is looking hard to stay out of the 15% bracket by then.  So what will doing the calculations about future RMDs help me understand?  Thanks for all the help!

So, if they are going to fall in the 15% with RMDs on mom's IRA, and will likely always do so, then the idea of converting to the top of the 15% bracket will allow those converted dollars to end up in Roth, which could never happen with the RMD money.Getting the conversions into Roth will reduce future RMDs which can never be converted.  Juggling that with the Medicare premium limits, social security taxation, etc. is the tricky part.  Should run some proforma on tax software to see the effects.Run the conversion numbers right up to the point where all the social security is being taxed (85%) and then look at the effective rates of conversion above that point to the point where Medicare premiums would go up, or the top of the bracket.  In the end, it really boils down to what mom/dad are willing to pay in taxes in order to convert to tax free Roth and their concern about future tax rates.

Also, remember that there is a "do over" available if too much is converted and taxes are too high or the conversion loses too much in value. A partial or full recharacterization of that conversion can be done up until 10/15 of the following year.

Throughout this evening, I assembled a spreadsheet to help me process various scenarios that I think we've been discussing. I assume it is far from perfect but I'm hoping that if I describe what I've done, someone here can tell me if I'm too far off and need to make adjustments before it is useful. What I've tried to do is peer out a few years to help me understand how much of the IRA, whether by conversion or by RMD, is being lost to taxes as described above. I assumed that Taxcaster for 2015 gets me close enough to make tax estimates for several years into the future. I used a tool from Schwab to compute RMDs for several years into the future too. I also assumed no COLA adjustments for the pension or the Social Security. I then generated the tax bill with and without monies removed from the IRA. I used the difference to understand what percent of the IRA monies are being sent to the IRS, right off the bat. I then made similar tabulations over a period of time to understand how much money came of the IRA and how much went to the IRS. I hope that made sense. If not, please ask for clarification and I'll try. But assuming I'm close, I'm seeing that the scenario where we never convert and only stick to RMDs for years, the least is lost to the IRS, but none ends up in the Roth IRA. In another extreme example, I converted them out to the Roth before mom's RMDs kick in. Slightly more is lost to the IRS than the other scenario, but within 2-3%, and the Roth IRA is grown significantly. Does that seem right? It is late and I've had a long day, so I could easily have made errors, but does anyone follow my description and have a thought?

I'm still trying to wrap my head around the analysis I need to complete. I thought I was making progress yesterday, but am needing to make several assumptions. Can someone verify these? Like, can Taxcaster get me close enough on tax bill estimates into future years (2016, 2017, 2018, etc.)? Can a 2015 RMD calculator get me close enough on estimates into future years (2016, 2017, 2018, etc.)? How many years of estimates should I include in my analysis (5, 10, 20, 40, etc.)? Is making an assumption to ignore COLA increases on the pension and benefits a reasonable one to make to keep calculations simple?


Find members of Ed Slott's Elite IRA Advisor GroupSM in your area.
We neither keep nor share your information entered on this form.

I agree to the terms and services:

You may review the terms and conditions here.