Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
User avatar
LadyGeek
Site Admin
Posts: 54913
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by LadyGeek » Mon Jun 03, 2019 8:20 pm

Lee_WSP wrote:
Mon Jun 03, 2019 7:23 pm
fyre4ce wrote:
Mon Jun 03, 2019 7:07 pm
LadyGeek wrote:
Mon Jun 03, 2019 6:01 pm
Bear in mind that the wiki supports mobile devices. I generally keep tables to 600px max.
For this article, what would be the method for limiting the width?
Limiting the width is a back end fix. Unless there's a layout option off to the side when you do your editing, it's not a single button click.
...
I'm using Chrome's Inspect to emulate a mobile device.

You can also emulate a mobile device by clicking on the "Mobile view" link at the very bottom of the page, then narrowing your browser window. to about 600px wide (or less). Click on "Desktop" at the very bottom of the page to switch back.

The first thing I notice is that the page is very long. Consider that a large portion of wiki readers are using a cellphone or tablet. They'll lose interest before hitting the end of the page, which can be counterproductive.

The next thing is that the math equations (which are actually images) are scrolling past the right-side border. Since they're images, they won't wrap to a new line like a text equation would.

I think we're close on a consensus (virtual fingers crossed) and all that's needed is some reorganization. Perhaps to incorporate Lee_WSP's suggestions by moving the examples to a new "User:Fyre4ce/Traditional versus Roth examples" page?
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.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Mon Jun 03, 2019 8:52 pm

FiveK wrote:
Mon Jun 03, 2019 5:38 pm
fyre4ce wrote:
Mon Jun 03, 2019 5:27 pm
I'm definitely willing to consider including it, but I think it needs some more words for context, and preferably the math should be converted to "pretty print" <math></math> display.
Good to know that you are willing to consider including it.

Yes, if others see no errors then it's worth taking the time to make it look better.
I think there may be some issues with your math. I tried to duplicate your results and I got different answers. I'm in a hurry but did some hand calcs, picture is here:

https://drive.google.com/file/d/19WQ1V7 ... sp=sharing

Please forgive my sloppy handwriting. I'm getting different answers for your take-home pay. You have a T^2 or R^2 terms and that doesn't look right.

I plugged in some simple numbers with a 22% marginal rate today and a 10% saver's credit that applies to both Roth and Traditional contributions. If the saver's credit didn't exist, normal advice would be to contribute Roth if future tax rates were >22%, Traditional if they were <22%, and a tie if they =22%. The Saver's credit appears to bias things slightly in favor of Traditional; Roth is only better now if rates are >24.44%, so even if future rates were 24% (2% higher than today) Traditional would still be better. The big mistake here would be to use a savings rate for Traditional that included the Saver's credit and ignore that you also get a Saver's credit for Roth.

I think the mechanism here is similar to that of an employee match. If you get a dollar-for-dollar match from your employer, favor Traditional because it's easier to run up a big number than with Roth. Likewise, it's easier to get a Saver's credit 10% "match" with Traditional, which is why the break-even future rate gets pushed up even with the credit is an equal percentage to both contributions.

Thoughts?

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Mon Jun 03, 2019 9:02 pm

fyre4ce wrote:
Mon Jun 03, 2019 8:52 pm
You have a T^2 or R^2 terms and that doesn't look right.
Yup, two typos when I transcribed my scratch notes. ;) Thanks!

Fortunately everything past those two typos was done on the correct equations - I think.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Mon Jun 03, 2019 9:07 pm

fyre4ce wrote:
Mon Jun 03, 2019 8:52 pm
I plugged in some simple numbers with a 22% marginal rate today and a 10% saver's credit that applies to both Roth and Traditional contributions.
One can't be in the 22% bracket and still qualify for a saver's credit.

It is possible to have a 22% marginal rate for traditional (12% bracket + 10% saver's credit) and 10% for Roth (just the saver's credit). That's the example used in the wiki.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Mon Jun 03, 2019 9:15 pm

fyre4ce wrote:
Mon Jun 03, 2019 8:52 pm
I plugged in some simple numbers with a 22% marginal rate today and a 10% saver's credit that applies to both Roth and Traditional contributions. If the saver's credit didn't exist, normal advice would be to contribute Roth if future tax rates were >22%, Traditional if they were <22%, and a tie if they =22%. The Saver's credit appears to bias things slightly in favor of Traditional; Roth is only better now if rates are >24.44%, so even if future rates were 24% (2% higher than today) Traditional would still be better. The big mistake here would be to use a savings rate for Traditional that included the Saver's credit and ignore that you also get a Saver's credit for Roth.
Looked closer at your example: the 24.4% should be compared to the 32% marginal rate for traditional (assuming that one could get a 10% saver's credit in the 22% bracket), so the 10% marginal rate for the Roth shifts the "breakeven" from 32% to 24.4%.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 04, 2019 9:06 am

FiveK wrote:
Mon Jun 03, 2019 9:02 pm
fyre4ce wrote:
Mon Jun 03, 2019 8:52 pm
You have a T^2 or R^2 terms and that doesn't look right.
Yup, two typos when I transcribed my scratch notes. ;) Thanks!

Fortunately everything past those two typos was done on the correct equations - I think.
Ah yes, ha, I was in too much of a hurry to notice the end results were mathematically identical. It might be another day or two but I can take a swing at cleaning it up.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 04, 2019 9:07 am

FiveK wrote:
Mon Jun 03, 2019 9:15 pm
fyre4ce wrote:
Mon Jun 03, 2019 8:52 pm
I plugged in some simple numbers with a 22% marginal rate today and a 10% saver's credit that applies to both Roth and Traditional contributions. If the saver's credit didn't exist, normal advice would be to contribute Roth if future tax rates were >22%, Traditional if they were <22%, and a tie if they =22%. The Saver's credit appears to bias things slightly in favor of Traditional; Roth is only better now if rates are >24.44%, so even if future rates were 24% (2% higher than today) Traditional would still be better. The big mistake here would be to use a savings rate for Traditional that included the Saver's credit and ignore that you also get a Saver's credit for Roth.
Looked closer at your example: the 24.4% should be compared to the 32% marginal rate for traditional (assuming that one could get a 10% saver's credit in the 22% bracket), so the 10% marginal rate for the Roth shifts the "breakeven" from 32% to 24.4%.
Yes, you're right; I think we're saying the same thing. If the saver's credit applies to both Traditional and Roth, the break-even is much closer to the traditional rate WITHOUT the saver's credit than with.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 04, 2019 11:27 am

nolesrule had a good observation in To Roth 401(k) or not - Bogleheads.org. Reproducing the reply below:

Good point. Chart below shows results for someone investing $25K/yr at 4% return, starting at age 22, then withdrawing 4%/yr in retirement.

Bars on the left show how much one would expect if contributing to and retiring at age 60. Bars on the right show the result if contributing to and retiring at age 55. The different colors show the amounts due to contributions at the various ages, starting at the bottom with blue for age 22, orange for age 23, etc.

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?

Image

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Tue Jun 04, 2019 11:35 am

FiveK wrote:
Tue Jun 04, 2019 11:27 am

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?
It confuses the difference between "actual cost of investing" with "invested dollars". A concept I learned on these forums.

Dollar for dollar, the Roth is better than Trad. Why? Because it grows tax free. However, putting in that same dollar into a Roth does not necessarily cost the same as that same dollar in Trad.

In the 24% bracket, that Roth dollar costs an actual $1.315, whereas the Trad dollar costs $1 due to taxes (tax drag). However, if one is in the lowest brackets, since taking the Trad deduction may not be beneficial or possible, the cost for each is the same (no tax drag).

And that's why this debate even exists.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 04, 2019 12:15 pm

Lee_WSP wrote:
Tue Jun 04, 2019 11:35 am
FiveK wrote:
Tue Jun 04, 2019 11:27 am

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?
It confuses the difference between "actual cost of investing" with "invested dollars".
nolesrule's point was "...we need a better visual analogy. The contributions going in year after year are like building a multi-layered cake."

This addresses the "marginal vs. effective" fallacy: because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.

The question is "does that graphic provide the better visual analogy (if one is needed)?"

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Tue Jun 04, 2019 12:36 pm

FiveK wrote:
Tue Jun 04, 2019 12:15 pm
Lee_WSP wrote:
Tue Jun 04, 2019 11:35 am
FiveK wrote:
Tue Jun 04, 2019 11:27 am

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?
It confuses the difference between "actual cost of investing" with "invested dollars".
nolesrule's point was "...we need a better visual analogy. The contributions going in year after year are like building a multi-layered cake."

This addresses the "marginal vs. effective" fallacy: because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.

The question is "does that graphic provide the better visual analogy (if one is needed)?"
Not sure I'm understanding the premise. Ie the cake analogy.

Money is fungible. It's not a cake. It's a pool.

What concept are we trying to convey with the visual analogy exactly?

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 04, 2019 1:56 pm

Lee_WSP wrote:
Tue Jun 04, 2019 12:36 pm
FiveK wrote:
Tue Jun 04, 2019 12:15 pm
Lee_WSP wrote:
Tue Jun 04, 2019 11:35 am
FiveK wrote:
Tue Jun 04, 2019 11:27 am

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?
It confuses the difference between "actual cost of investing" with "invested dollars".
nolesrule's point was "...we need a better visual analogy. The contributions going in year after year are like building a multi-layered cake."

This addresses the "marginal vs. effective" fallacy: because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.

The question is "does that graphic provide the better visual analogy (if one is needed)?"
Not sure I'm understanding the premise. Ie the cake analogy.

Money is fungible. It's not a cake. It's a pool.

What concept are we trying to convey with the visual analogy exactly?
Because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Tue Jun 04, 2019 1:59 pm

FiveK wrote:
Tue Jun 04, 2019 1:56 pm
Lee_WSP wrote:
Tue Jun 04, 2019 12:36 pm
FiveK wrote:
Tue Jun 04, 2019 12:15 pm
Lee_WSP wrote:
Tue Jun 04, 2019 11:35 am
FiveK wrote:
Tue Jun 04, 2019 11:27 am

Any thoughts on whether this (or something similar) would be useful in the Traditional vs. Roth wiki?
It confuses the difference between "actual cost of investing" with "invested dollars".
nolesrule's point was "...we need a better visual analogy. The contributions going in year after year are like building a multi-layered cake."

This addresses the "marginal vs. effective" fallacy: because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.

The question is "does that graphic provide the better visual analogy (if one is needed)?"
Not sure I'm understanding the premise. Ie the cake analogy.

Money is fungible. It's not a cake. It's a pool.

What concept are we trying to convey with the visual analogy exactly?
Because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.
That's what I thought you were saying. But that's not how it works. Retirement accounts don't have that tax basis component. The money is all taxed or not taxed at the same rate upon withdrawal (trad) or contribution (Roth). Once in the trad or Roth account, it's all fungible & mixed in.

In reality, the money in brokerage accounts are also technically mixed in, but by using accounting, we can calculate how much each individual transaction lost or gained relative to the pot as a whole.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 04, 2019 2:46 pm

Lee_WSP wrote:
Tue Jun 04, 2019 1:59 pm
FiveK wrote:
Tue Jun 04, 2019 1:56 pm
Lee_WSP wrote:
Tue Jun 04, 2019 12:36 pm
Not sure I'm understanding the premise. Ie the cake analogy.
Money is fungible. It's not a cake. It's a pool.
What concept are we trying to convey with the visual analogy exactly?
Because withdrawals based on each subsequent year's contribution are taken "on top of" withdrawals based on all previous years' contributions, each subsequent year's withdrawals are taxed at a marginal, not effective, rate.
That's what I thought you were saying. But that's not how it works. Retirement accounts don't have that tax basis component. The money is all taxed or not taxed at the same rate upon withdrawal (trad) or contribution (Roth). Once in the trad or Roth account, it's all fungible & mixed in.

In reality, the money in brokerage accounts are also technically mixed in, but by using accounting, we can calculate how much each individual transaction lost or gained relative to the pot as a whole.
But it is how "it" works, provided we define "it" appropriately.

Take a 50 year old who already has a substantial traditional balance -say, $500K. Even without any further contributions, that could reasonably double to $1 million by age 65. Taking a 4%/yr withdrawal then gives $40K/yr.

"It" is the decision being made at age 51 whether to contribute to traditional or Roth. Any traditional contributions at age 51 will increase the traditional balance at age 65, thus allowing more that $40K/yr withdrawal. The taxation on the amount above $40K/yr will occur at the marginal rate on that amount, not the effective rate on the total income.

Does that explain "it" well enough?

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Tue Jun 04, 2019 2:56 pm

I see what your saying now.

I still don't think the cake analogy is correct since what you're doing is increasing the size of the cake rather than adding on another layer. Although the layer visual can adequately depict the brackets. Sort of like volume markers in a beaker.

edit:
Additionally, the marginal rate on withdrawals will not change if you maintain the same exact level of withdrawals. Just because you increase the size of the retirement portfolio does not mean you have to or will take out any more money than planned.

nolesrule
Posts: 1246
Joined: Thu Feb 26, 2015 10:59 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by nolesrule » Tue Jun 04, 2019 8:35 pm

Lee_WSP wrote:
Tue Jun 04, 2019 2:56 pm
I see what your saying now.

I still don't think the cake analogy is correct since what you're doing is increasing the size of the cake rather than adding on another layer. Although the layer visual can adequately depict the brackets. Sort of like volume markers in a beaker.
You are increasing the size of the cake by adding another layer if you make a Traditional contribution. If you make a Roth contribution, then no layer gets added to the cake and ultimately the cake is smaller.

So for example, you are in the 22% bracket now, all of the this year's contribution is 22%. You look at the size of your cake currently and also with the additional layer of the potential contribution added in and take a slice of each cake (withdrawal) and use that to estimate your future taxes. Without that layer the top rate on your withdrawal may be 12%. But by adding another layer, maybe the top rate is 22%. That's the decision point one is trying to make with each year's contributions.
edit:
Additionally, the marginal rate on withdrawals will not change if you maintain the same exact level of withdrawals. Just because you increase the size of the retirement portfolio does not mean you have to or will take out any more money than planned.
Of course not, but you need to make some assumptions or you can't do any type of analysis.

The biggest problem I see is that too many people think "Yeah, I'll be in a higher tax bracket in retirement, because I make a lot of money and contribute the max, so I'll go Roth", making the wrong decision for one reason or another because they didn't work long enough.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 04, 2019 9:16 pm

Lee_WSP wrote:
Tue Jun 04, 2019 2:56 pm
I see what your saying now.
Ok, thanks. I changed the example in the wiki to use the explanation given here. Never have been completely happy about that old explanation....

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Tue Jun 04, 2019 10:33 pm

nolesrule wrote:
Tue Jun 04, 2019 8:35 pm
Lee_WSP wrote:
Tue Jun 04, 2019 2:56 pm
I see what your saying now.

I still don't think the cake analogy is correct since what you're doing is increasing the size of the cake rather than adding on another layer. Although the layer visual can adequately depict the brackets. Sort of like volume markers in a beaker.
You are increasing the size of the cake by adding another layer if you make a Traditional contribution. If you make a Roth contribution, then no layer gets added to the cake and ultimately the cake is smaller.

So for example, you are in the 22% bracket now, all of the this year's contribution is 22%. You look at the size of your cake currently and also with the additional layer of the potential contribution added in and take a slice of each cake (withdrawal) and use that to estimate your future taxes. Without that layer the top rate on your withdrawal may be 12%. But by adding another layer, maybe the top rate is 22%. That's the decision point one is trying to make with each year's contributions.
I'm not following this logic.

nolesrule
Posts: 1246
Joined: Thu Feb 26, 2015 10:59 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by nolesrule » Wed Jun 05, 2019 7:19 am

Lee_WSP wrote:
Tue Jun 04, 2019 10:33 pm
nolesrule wrote:
Tue Jun 04, 2019 8:35 pm
Lee_WSP wrote:
Tue Jun 04, 2019 2:56 pm
I see what your saying now.

I still don't think the cake analogy is correct since what you're doing is increasing the size of the cake rather than adding on another layer. Although the layer visual can adequately depict the brackets. Sort of like volume markers in a beaker.
You are increasing the size of the cake by adding another layer if you make a Traditional contribution. If you make a Roth contribution, then no layer gets added to the cake and ultimately the cake is smaller.

So for example, you are in the 22% bracket now, all of the this year's contribution is 22%. You look at the size of your cake currently and also with the additional layer of the potential contribution added in and take a slice of each cake (withdrawal) and use that to estimate your future taxes. Without that layer the top rate on your withdrawal may be 12%. But by adding another layer, maybe the top rate is 22%. That's the decision point one is trying to make with each year's contributions.
I'm not following this logic.
That's why I think we need a visual.

Previous contributions effect the withdrawal rate on future contributions, but not vice versa. That's where the layers come in. And because most people don't pull everything out in one withdrawal, that's why the withdrawal is a slice of all the layers. You have to look at where the slices line up with the tax brackets when income stacking.

Let's say for example you contribute $10k to Traditional for 20 years (I'm ignoring growth for simple illustration). You have $200k. What's the tax rate going to be upon withdrawal in retirement? You're not withdrawing all at once. Let's say you withdraw 4% in retirement. Your withdrawal is $8k.

Now, in year 21 you want to add another $10k. If you add to traditional, your 4% withdrawal in retirement becomes $8.4k. It's the tax rate on that $400 that you are comparing to the marginal rate on the $10k going in. The other $8k is only relevant to the decision because the $400 sits on top of it in the income stack for determining the in and out rates on the contribution.

The dollars may be fungible, but the effects of the contribution decision of those dollars is not fungible.

rkhusky
Posts: 6753
Joined: Thu Aug 18, 2011 8:09 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by rkhusky » Wed Jun 05, 2019 7:26 am

Lee_WSP wrote:
Mon Jun 03, 2019 12:22 pm
rkhusky wrote:
Mon Jun 03, 2019 8:35 am
Lee_WSP wrote:
Sun Jun 02, 2019 10:05 pm
rkhusky wrote:
Sun Jun 02, 2019 9:21 pm
Lee_WSP wrote:
Sun Jun 02, 2019 7:31 pm
Sure. Tell give me the estimated hump in your situation and what you're choices are. We can work through it together.
$20K pension, $30K SS, $60K Roth/Traditional IRA withdrawal. Contribution tax rate is 15%.
And this hypothetical is trying to figure out how to allocate what/how much ?
MFJ.

Roth/Traditional split for contributions at 15% contribution rate.

Roth/Traditional split for contributions at 22% contribution rate.

Roth/Traditional split for withdrawals.
Easy. You contribute up to the maximum ceiling of the 15% contribution bracket to a Roth. The rest in Trad.

Then during retirement, you alternate your withdrawals as follows:
1 year all Trad + Roth conversions up to the bracket ceiling.
Next year all Roth up to the 60k living expenses per fact pattern.

Repeat until Trad account is depleted or you're facing RMD's.

edit: Or even more efficiently:

DAlternate your withdrawals as follows:
1 year all Trad + Roth conversions up to the bracket ceiling.
Next year Trad up to the SS cliff the rest Roth up to the 60k living expenses.
So, you don't recommend actually looking at your tax rates when planning for retirement withdrawals?

Using the MFJ heat map in https://www.bogleheads.org/wiki/Taxatio ... y_benefits, it looks like, for the numbers in the example, that the IRA withdrawals will pass through 0%, 15%, 18.5%, 22.2% and 12% marginal rates. If the withdrawals are all from Traditional, then the average tax rate on the IRA withdrawals is about 14.8%. This would imply that one should contribute fully to Traditional, whether in 15% or 22% tax brackets on contribution (although using Roth in 15% bracket is not much of an error).

Decreasing pension and increasing IRA withdrawal, while keeping the total at $80K, increases the benefit of using Traditional in the 15% bracket.
Reducing the Traditional IRA withdrawal (by using some Roth or by reducing IRA withdrawal), but keeping the pension the same, increases the benefit of using Roth in the 15% bracket.

If instead, SS was $20K, $20K pension, $68K IRA withdrawal, the average tax rate on the Traditional IRA withdrawal would be 13.1%, which more strongly suggests Traditional contributions in the 15% contribution tax bracket.

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Wed Jun 05, 2019 10:29 am

@Rhusky
The only firm conclusion you can come to with your example is the same one in the wiki. You should contribute to both. You should have a big pool to draw from from both Roth & Trad to be able to be flexible during withdrawals.

Chasing 1% tax savings that may or may change with a new tax law or revision of guidance is not terribly helpful in the present.

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Wed Jun 05, 2019 10:47 am

nolesrule wrote:
Wed Jun 05, 2019 7:19 am
Lee_WSP wrote:
Tue Jun 04, 2019 10:33 pm
nolesrule wrote:
Tue Jun 04, 2019 8:35 pm
Lee_WSP wrote:
Tue Jun 04, 2019 2:56 pm
I see what your saying now.

I still don't think the cake analogy is correct since what you're doing is increasing the size of the cake rather than adding on another layer. Although the layer visual can adequately depict the brackets. Sort of like volume markers in a beaker.
You are increasing the size of the cake by adding another layer if you make a Traditional contribution. If you make a Roth contribution, then no layer gets added to the cake and ultimately the cake is smaller.

So for example, you are in the 22% bracket now, all of the this year's contribution is 22%. You look at the size of your cake currently and also with the additional layer of the potential contribution added in and take a slice of each cake (withdrawal) and use that to estimate your future taxes. Without that layer the top rate on your withdrawal may be 12%. But by adding another layer, maybe the top rate is 22%. That's the decision point one is trying to make with each year's contributions.
I'm not following this logic.
That's why I think we need a visual.

Previous contributions effect the withdrawal rate on future contributions, but not vice versa. That's where the layers come in. And because most people don't pull everything out in one withdrawal, that's why the withdrawal is a slice of all the layers. You have to look at where the slices line up with the tax brackets when income stacking.

Let's say for example you contribute $10k to Traditional for 20 years (I'm ignoring growth for simple illustration). You have $200k. What's the tax rate going to be upon withdrawal in retirement? You're not withdrawing all at once. Let's say you withdraw 4% in retirement. Your withdrawal is $8k.

Now, in year 21 you want to add another $10k. If you add to traditional, your 4% withdrawal in retirement becomes $8.4k. It's the tax rate on that $400 that you are comparing to the marginal rate on the $10k going in. The other $8k is only relevant to the decision because the $400 sits on top of it in the income stack for determining the in and out rates on the contribution.

The dollars may be fungible, but the effects of the contribution decision of those dollars is not fungible.
Let me try and understand the parameters:
  • There will be X% withdrawal rate off the total retirement pool
  • We're going to ignore Roth for the purposes of this illustration
  • We're trying to illustrate how much current contributions will be taxed upon withdrawal
If I've got that right, let me propose the following.

I still disagree that it is a cake because the money is fungible and there is no specific basis with retirement accounts. The government currently does not care when you deposited it or what the basis is. So, it's more of a homogenous mix.

I propose thinking of it like a balloon which will expel X% each time period. As the balloon gets bigger, the total amount exiting each time period gets bigger as it is a portion of the total amount in the balloon.

So, each time you put more air into the balloon, the balloon gets bigger and the amount exiting gets bigger.

Knowing this, we can then surmise that by putting more air into the balloon will necessitate a larger expelling upon deflation time. However, the total amount expelled per period will continually decrease as the balloon shrinks in size (assuming no growth or at least assume withdrawals > growth).

The effect of putting in more trad contributions thus becomes:
You will increase the total amount of withdrawals each year as you will be withdrawing x%, however, if the percentage of withdrawals is > % growth, your amount of withdrawals per year will decrease over time. Likewise if %withdrawals < %growth your withdrawals will increase with time up until RMD's kick in.

If the trad account is at the margin & one more contribution will push it over into the next tax bracket, we can therefore say:
Some of this contribution will be taxed at the higher rate, but not all of it. And only a small portion will be taxed at the higher rate during retirement because that's how tax brackets work. In the next year, it is entirely possible and probable that the total withdrawals will be below the bracket margin and thus the contribution will not be taxed at a higher bracket.

But saying that that particular contribution ignores the fact that money is fungible, but for the purposes of this thought experiment, let's just say you can track it via specific basis. It's more accurate to say that Y% of the withdrawal (regardless of where the original contribution came from) will be taxed at the higher rate.

rkhusky
Posts: 6753
Joined: Thu Aug 18, 2011 8:09 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by rkhusky » Wed Jun 05, 2019 11:12 am

Lee_WSP wrote:
Wed Jun 05, 2019 10:29 am
@Rhusky
The only firm conclusion you can come to with your example is the same one in the wiki. You should contribute to both. You should have a big pool to draw from from both Roth & Trad to be able to be flexible during withdrawals.

Chasing 1% tax savings that may or may change with a new tax law or revision of guidance is not terribly helpful in the present.
The conclusion that I draw is that one should run the numbers to see what one should do, rather than blindly follow some ad hoc rule (e.g. withdrawing from Roth with a 0% marginal rate). And I'm not seeing a SS cliff.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Wed Jun 05, 2019 11:49 am

LadyGeek wrote:
Mon Jun 03, 2019 8:20 pm
The first thing I notice is that the page is very long. Consider that a large portion of wiki readers are using a cellphone or tablet. They'll lose interest before hitting the end of the page, which can be counterproductive.

The next thing is that the math equations (which are actually images) are scrolling past the right-side border. Since they're images, they won't wrap to a new line like a text equation would.

I think we're close on a consensus (virtual fingers crossed) and all that's needed is some reorganization. Perhaps to incorporate Lee_WSP's suggestions by moving the examples to a new "User:Fyre4ce/Traditional versus Roth examples" page?
I like the idea of an examples page.

Started some reorganization to highlight the main points up front.

User avatar
Kevin M
Posts: 10883
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Kevin M » Wed Jun 05, 2019 1:37 pm

nolesrule wrote:
Wed Jun 05, 2019 7:19 am
That's why I think we need a visual.
Visuals work better for some people than others.
nolesrule wrote:
Wed Jun 05, 2019 7:19 am
Previous contributions effect the withdrawal rate on future contributions, but not vice versa. That's where the layers come in. And because most people don't pull everything out in one withdrawal, that's why the withdrawal is a slice of all the layers. You have to look at where the slices line up with the tax brackets when income stacking.
I didn't understand this too well.
nolesrule wrote:
Wed Jun 05, 2019 7:19 am
Let's say for example you contribute $10k to Traditional for 20 years (I'm ignoring growth for simple illustration). You have $200k. What's the tax rate going to be upon withdrawal in retirement? You're not withdrawing all at once. Let's say you withdraw 4% in retirement. Your withdrawal is $8k.

Now, in year 21 you want to add another $10k. If you add to traditional, your 4% withdrawal in retirement becomes $8.4k. It's the tax rate on that $400 that you are comparing to the marginal rate on the $10k going in. The other $8k is only relevant to the decision because the $400 sits on top of it in the income stack for determining the in and out rates on the contribution.
I found this to be the most understandable example so far, especially the underlined part, but maybe I didn't read previous attempts carefully enough.

The cake visual made no sense to me with a quick look. Your explanation makes a lot of sense with a quick read.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Wed Jun 05, 2019 3:40 pm

Kevin M wrote:
Wed Jun 05, 2019 1:37 pm
The cake visual made no sense to me with a quick look. Your explanation makes a lot of sense with a quick read.

Kevin
What about my attempt to re-envision it as a balloon? I couldn't think of anything else that would both expand when filled further and give off more during withdrawals.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Wed Jun 05, 2019 4:20 pm

FiveK wrote:
Wed Jun 05, 2019 11:49 am
LadyGeek wrote:
Mon Jun 03, 2019 8:20 pm
The first thing I notice is that the page is very long. Consider that a large portion of wiki readers are using a cellphone or tablet. They'll lose interest before hitting the end of the page, which can be counterproductive.

The next thing is that the math equations (which are actually images) are scrolling past the right-side border. Since they're images, they won't wrap to a new line like a text equation would.

I think we're close on a consensus (virtual fingers crossed) and all that's needed is some reorganization. Perhaps to incorporate Lee_WSP's suggestions by moving the examples to a new "User:Fyre4ce/Traditional versus Roth examples" page?
I like the idea of an examples page.

Started some reorganization to highlight the main points up front.
Apologies if this jams things up, but my opinion is that examples are more effective when they immediately follow the theory they are trying to help explain. I worry that if we segregate all the examples on a separate page, fewer readers will be able to absorb the content. May I propose the following alternatives for reorganization:

1) Create an "advanced" or "complex" section at the bottom that includes: taxation of SS, splitting brackets, maxing out accounts, irregular rates, saver's credit
2) Split the page into a "basic" and "complex"/"advanced" pages, and move the advanced content onto the advanced page, with in-line examples

In the "basic" section, there are only two examples: the most basic one showing how marginal tax rates affect after-tax value, and an example showing how to estimate future tax rates. In my opinion these are helpful enough for readers to justify their inclusion toward the top of the article.

Lee_WSP
Posts: 391
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by Lee_WSP » Wed Jun 05, 2019 5:00 pm

There may be a "click to expand" option.

nolesrule
Posts: 1246
Joined: Thu Feb 26, 2015 10:59 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by nolesrule » Wed Jun 05, 2019 5:03 pm

Kevin M wrote:
Wed Jun 05, 2019 1:37 pm
nolesrule wrote:
Wed Jun 05, 2019 7:19 am
That's why I think we need a visual.
Visuals work better for some people than others.
nolesrule wrote:
Wed Jun 05, 2019 7:19 am
Previous contributions effect the withdrawal rate on future contributions, but not vice versa. That's where the layers come in. And because most people don't pull everything out in one withdrawal, that's why the withdrawal is a slice of all the layers. You have to look at where the slices line up with the tax brackets when income stacking.
I didn't understand this too well.
nolesrule wrote:
Wed Jun 05, 2019 7:19 am
Let's say for example you contribute $10k to Traditional for 20 years (I'm ignoring growth for simple illustration). You have $200k. What's the tax rate going to be upon withdrawal in retirement? You're not withdrawing all at once. Let's say you withdraw 4% in retirement. Your withdrawal is $8k.

Now, in year 21 you want to add another $10k. If you add to traditional, your 4% withdrawal in retirement becomes $8.4k. It's the tax rate on that $400 that you are comparing to the marginal rate on the $10k going in. The other $8k is only relevant to the decision because the $400 sits on top of it in the income stack for determining the in and out rates on the contribution.
I found this to be the most understandable example so far, especially the underlined part, but maybe I didn't read previous attempts carefully enough.

The cake visual made no sense to me with a quick look. Your explanation makes a lot of sense with a quick read.

Kevin
Well, to be fair I came up with the cake idea while on medication that seems to have impaired my mental faculties somewhat over last few days. That's finally starting to clear up this afternoon (and I hope to never have to take that stuff again or if I have to, just refrain from posting). Combine that with the fact that I'm not exactly the most eloquent to begin with and I can understand the confusion.

But the cake analogy was inspired by Harry Sit's income and progressive tax system graphic from this article on the Case against Roth 401k. Showing the individual strata of each contribution's component of the withdrawals in that context is what I was trying to get at. It is somewhat simplistic because of the SS hump, but I think it gets the general idea across for those that need a visual.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Wed Jun 05, 2019 9:02 pm

fyre4ce wrote:
Wed Jun 05, 2019 4:20 pm
FiveK wrote:
Wed Jun 05, 2019 11:49 am
LadyGeek wrote:
Mon Jun 03, 2019 8:20 pm
The first thing I notice is that the page is very long. Consider that a large portion of wiki readers are using a cellphone or tablet. They'll lose interest before hitting the end of the page, which can be counterproductive.

The next thing is that the math equations (which are actually images) are scrolling past the right-side border. Since they're images, they won't wrap to a new line like a text equation would.

I think we're close on a consensus (virtual fingers crossed) and all that's needed is some reorganization. Perhaps to incorporate Lee_WSP's suggestions by moving the examples to a new "User:Fyre4ce/Traditional versus Roth examples" page?
I like the idea of an examples page.

Started some reorganization to highlight the main points up front.
Apologies if this jams things up, but my opinion is that examples are more effective when they immediately follow the theory they are trying to help explain. I worry that if we segregate all the examples on a separate page, fewer readers will be able to absorb the content. May I propose the following alternatives for reorganization:

1) Create an "advanced" or "complex" section at the bottom that includes: taxation of SS, splitting brackets, maxing out accounts, irregular rates, saver's credit
2) Split the page into a "basic" and "complex"/"advanced" pages, and move the advanced content onto the advanced page, with in-line examples

In the "basic" section, there are only two examples: the most basic one showing how marginal tax rates affect after-tax value, and an example showing how to estimate future tax rates. In my opinion these are helpful enough for readers to justify their inclusion toward the top of the article.
As someone accessing the forum via a desktop PC, page width and article length aren't constraints.

I defer to those more familiar with cellphone and tablet access to inform best practice there.

Draft page User:FiveK/Traditional versus Roth examples has been created. All wiki editors are free to modify per the usual guidelines. Whether that becomes a separate page in the live wiki, or examples are interwoven with the main Traditional vs. Roth article, matters not to me.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Fri Jun 07, 2019 8:54 pm

Changed the name of the "Marginal tax rates" section to "Two common misconceptions". Reordered some of the text and simplified others, all with the intent of putting summary information earlier to the reader, and either using references or later text for more details.

User avatar
LadyGeek
Site Admin
Posts: 54913
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by LadyGeek » Fri Jun 07, 2019 10:43 pm

^^^ That's for User:Fyre4ce/Traditional versus Roth.

This post has some discussion to change the section title "Irregular marginal rates" to "Real-world example". "Irregular" is not understood in the intended context, i.e. what is an "irregular" vs. a "regular" marginal rate?
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
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Fri Jun 07, 2019 11:25 pm

LadyGeek wrote:
Fri Jun 07, 2019 10:43 pm
^^^ That's for User:Fyre4ce/Traditional versus Roth.

This post has some discussion to change the section title "Irregular marginal rates" to "Real-world example". "Irregular" is not understood in the intended context, i.e. what is an "irregular" vs. a "regular" marginal rate?
Done. But there are more examples that are also "real world" in the article, so we need either to include this one with those under a "Real-world examples" section, move all (most?) of the examples to the "examples" article, or...?

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 11, 2019 1:01 pm

Trying to keep the ball moving forward with the edits.

We still do not have unanimous agreement on the appropriate definition of marginal rate. I still feel, strongly, that we should stick to the commonly used "small" definition as opposed to coming up with our own. I think the "no original research" policy supports the idea of using the most common definitions. I also happen to disagree that using the "any amount" definition makes things easier to understand. As I explained in my banana buying example, there's a difference between the unit price and average unit cost, and using the "any amount" definition confuses these two things. I know FiveK has been a supporter of the "any amount" definition, and while I respect the contributions he's made to the forum and wiki (which are far greater than mine), I still disagree with him on this particular point. I think we completely agree on the overall analysis methodology, just disagree on the best language to use. We've spilled more ink on this definition question than on any other aspect of these pages, which is unfortunate. Nonetheless, I don't want to strong-arm a path that will leave some contributors unhappy. In the spirit of moving things along I have a few questions:

-Is there anyone other than FiveK who supports the "any amount" definition of marginal tax rate, as opposed to "small"?
-What suggestions do people have for working to resolve this disagreement? Vote? Compromise? (eg. "The conventional definition is ____, however in certain situations it may be helpful to use an alternate definition ______")
-For those who support the "any amount" definition, in irregular rate situations where the rate is being taken over a larger contribution, what term do you propose for the "small" change tax rate that I would prefer to call marginal? This value is produced by the personal finance toolbox plots and is still a real thing; if we call the average contribution tax rate a "marginal tax rate" then we'll need a new term for the tax rate for small changes.

Sorry, but I don't agree with some of the edits that have been made recently. I think the recommendation of "Traditional during peak earning years, Roth otherwise" is about as compact and high-yield as it gets, and should be placed in the general guidelines. If people thought that section was too long, I think some of the content could be moved out (for example, comments about employer plans could be moved down to to Investment Choices , or removed entirely), but I think the general guidelines was better before than it is now. I'm OK adding that if the decision is close, split 50/50 for tax diversification.

I don't agree that investment performance should be estimated conservatively in a T vs R decision; it should be invested as accurately as possible (neither conservatively nor liberally). For example, if I told you I'd pay you $10,000 if you correctly predicted your account balances on your 65th birthday within 5%, would you use conservative estimates or accurate estimates? If you currently hold all bonds with a yield of 3.5%, would you assume 3.5% growth or 2.5% growth? I'd assume 3.5% if I wanted to win the bet; rates may go up or down, but assuming current rates gives me the best shot. In some situations, like estimating how much you need to save for retirement, being conservative makes sense, because having too little money is very bad, but having too much is not a problem. In the T vs R analysis, it's more symmetric - if growth is higher than estimated, you pay more taxes later, and if growth is less than expected, you pay more taxes now. I don't think conservative is good here; it will, on average, cause you to pay too much tax in retirement.

I'd like to add a few sections in the Marginal Rate special cases section for rates <0 and >100%. I think it will help people understand how rates work. Can anyone provide a real-world example of either case?

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 2:35 pm

When the personal finance community switches from using "marginal" to using "average marginal" to describe what tax rates should be compared on the contribution and withdrawal ends, then we can revert to using the calculus-based definition of marginal.

Until then we should help people by explaining marginal in a way that applies to the non-smooth realm of personal finance and the tax code. Referencing Kitces' article about "the amount of income involved" serves admirably.

It has been a bit disappointing to find that the Bogleheads' Guide to Retirement Planning may be responsible for some of the "marginal vs. effective" misunderstanding, but nobody's perfect. ;)

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 11, 2019 2:53 pm

FiveK wrote:
Tue Jun 11, 2019 2:35 pm
When the personal finance community switches from using "marginal" to using "average marginal" to describe what tax rates should be compared on the contribution and withdrawal ends, then we can revert to using the calculus-based definition of marginal.

Until then we should help people by explaining marginal in a way that applies to the non-smooth realm of personal finance and the tax code. Referencing Kitces' article about "the amount of income involved" serves admirably.

It has been a bit disappointing to find that the Bogleheads' Guide to Retirement Planning may be responsible for some of the "marginal vs. effective" misunderstanding, but nobody's perfect. ;)
How do you suggest we move forward?

trueblueky
Posts: 1509
Joined: Tue May 27, 2014 3:50 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by trueblueky » Tue Jun 11, 2019 3:28 pm

fyre4ce wrote:
Tue Jun 11, 2019 1:01 pm
Trying to keep the ball moving forward with the edits.

We still do not have unanimous agreement on the appropriate definition of marginal rate. I still feel, strongly, that we should stick to the commonly used "small" definition as opposed to coming up with our own. I think the "no original research" policy supports the idea of using the most common definitions. I also happen to disagree that using the "any amount" definition makes things easier to understand. As I explained in my banana buying example, there's a difference between the unit price and average unit cost, and using the "any amount" definition confuses these two things. I know FiveK has been a supporter of the "any amount" definition, and while I respect the contributions he's made to the forum and wiki (which are far greater than mine), I still disagree with him on this particular point. I think we completely agree on the overall analysis methodology, just disagree on the best language to use. We've spilled more ink on this definition question than on any other aspect of these pages, which is unfortunate. Nonetheless, I don't want to strong-arm a path that will leave some contributors unhappy. In the spirit of moving things along I have a few questions:

-Is there anyone other than FiveK who supports the "any amount" definition of marginal tax rate, as opposed to "small"?
-What suggestions do people have for working to resolve this disagreement? Vote? Compromise? (eg. "The conventional definition is ____, however in certain situations it may be helpful to use an alternate definition ______")
-For those who support the "any amount" definition, in irregular rate situations where the rate is being taken over a larger contribution, what term do you propose for the "small" change tax rate that I would prefer to call marginal? This value is produced by the personal finance toolbox plots and is still a real thing; if we call the average contribution tax rate a "marginal tax rate" then we'll need a new term for the tax rate for small changes.

Sorry, but I don't agree with some of the edits that have been made recently. I think the recommendation of "Traditional during peak earning years, Roth otherwise" is about as compact and high-yield as it gets, and should be placed in the general guidelines. If people thought that section was too long, I think some of the content could be moved out (for example, comments about employer plans could be moved down to to Investment Choices , or removed entirely), but I think the general guidelines was better before than it is now. I'm OK adding that if the decision is close, split 50/50 for tax diversification.

I don't agree that investment performance should be estimated conservatively in a T vs R decision; it should be invested as accurately as possible (neither conservatively nor liberally). For example, if I told you I'd pay you $10,000 if you correctly predicted your account balances on your 65th birthday within 5%, would you use conservative estimates or accurate estimates? If you currently hold all bonds with a yield of 3.5%, would you assume 3.5% growth or 2.5% growth? I'd assume 3.5% if I wanted to win the bet; rates may go up or down, but assuming current rates gives me the best shot. In some situations, like estimating how much you need to save for retirement, being conservative makes sense, because having too little money is very bad, but having too much is not a problem. In the T vs R analysis, it's more symmetric - if growth is higher than estimated, you pay more taxes later, and if growth is less than expected, you pay more taxes now. I don't think conservative is good here; it will, on average, cause you to pay too much tax in retirement.

I'd like to add a few sections in the Marginal Rate special cases section for rates <0 and >100%. I think it will help people understand how rates work. Can anyone provide a real-world example of either case?
Real-world marginal rate > 100%
MFJ no children, AGI $38,000 reflects $2000 contribution to IRA. Tax before credit is $1403 per tax table. Couple receives savers credit of $1000. Tax = $403.
MFJ no children, AGI $38,001 reflects $2000 contribution to IRA. Tax before credit is still $1403. Couple receives savers credit of $400. Tax = $1003
One dollar increase in income yields $600 additional tax.

Real.world marginal rate <0%
MFJ 3 children, AGI $14,249. No tax due before applying credits. EITC = $6401
MFJ 3 children, AGI $14,250. No tax due before applying credits. EITC = $6431
One dollar more in income yields $30 more in refund.

If you include the end of the ACA subsidy at 400% of the poverty rate, one more dollar can mean $1000s more owed with tax return.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 3:33 pm

fyre4ce wrote:
Tue Jun 11, 2019 2:53 pm
FiveK wrote:
Tue Jun 11, 2019 2:35 pm
When the personal finance community switches from using "marginal" to using "average marginal" to describe what tax rates should be compared on the contribution and withdrawal ends, then we can revert to using the calculus-based definition of marginal.

Until then we should help people by explaining marginal in a way that applies to the non-smooth realm of personal finance and the tax code. Referencing Kitces' article about "the amount of income involved" serves admirably.

It has been a bit disappointing to find that the Bogleheads' Guide to Retirement Planning may be responsible for some of the "marginal vs. effective" misunderstanding, but nobody's perfect. ;)
How do you suggest we move forward?
Taking the draft versions as of this date and time and moving them to the public wiki would work.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 11, 2019 3:45 pm

trueblueky wrote:
Tue Jun 11, 2019 3:28 pm
Real-world marginal rate > 100%
MFJ no children, AGI $38,000 reflects $2000 contribution to IRA. Tax before credit is $1403 per tax table. Couple receives savers credit of $1000. Tax = $403.
MFJ no children, AGI $38,001 reflects $2000 contribution to IRA. Tax before credit is still $1403. Couple receives savers credit of $400. Tax = $1003
One dollar increase in income yields $600 additional tax.

Real.world marginal rate <0%
MFJ 3 children, AGI $14,249. No tax due before applying credits. EITC = $6401
MFJ 3 children, AGI $14,250. No tax due before applying credits. EITC = $6431
One dollar more in income yields $30 more in refund.

If you include the end of the ACA subsidy at 400% of the poverty rate, one more dollar can mean $1000s more owed with tax return.
Thanks! Are there any examples of <0 or >100% marginal rates over a significant range, as opposed to a single spike?

rkhusky
Posts: 6753
Joined: Thu Aug 18, 2011 8:09 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by rkhusky » Tue Jun 11, 2019 3:50 pm

fyre4ce wrote:
Tue Jun 11, 2019 2:53 pm
How do you suggest we move forward?
I added some references for the common usage of next/last dollar, but they seem to have been deleted. I was going to suggest that someone add a sentence about other usages for marginal, and associated references.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 3:54 pm

rkhusky wrote:
Tue Jun 11, 2019 3:50 pm
fyre4ce wrote:
Tue Jun 11, 2019 2:53 pm
How do you suggest we move forward?
I added some references for the common usage of next/last dollar, but they seem to have been deleted. I was going to suggest that someone add a sentence about other usages for marginal, and associated references.
Moved, not deleted (I believe). At least, the intent was "move".

trueblueky
Posts: 1509
Joined: Tue May 27, 2014 3:50 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by trueblueky » Tue Jun 11, 2019 3:59 pm

fyre4ce wrote:
Tue Jun 11, 2019 3:45 pm
trueblueky wrote:
Tue Jun 11, 2019 3:28 pm
Real-world marginal rate > 100%
MFJ no children, AGI $38,000 reflects $2000 contribution to IRA. Tax before credit is $1403 per tax table. Couple receives savers credit of $1000. Tax = $403.
MFJ no children, AGI $38,001 reflects $2000 contribution to IRA. Tax before credit is still $1403. Couple receives savers credit of $400. Tax = $1003
One dollar increase in income yields $600 additional tax.

Real.world marginal rate <0%
MFJ 3 children, AGI $14,249. No tax due before applying credits. EITC = $6401
MFJ 3 children, AGI $14,250. No tax due before applying credits. EITC = $6431
One dollar more in income yields $30 more in refund.

If you include the end of the ACA subsidy at 400% of the poverty rate, one more dollar can mean $1000s more owed with tax return.
Thanks! Are there any examples of <0 or >100% marginal rates over a significant range, as opposed to a single spike?
The EITC has < 0% all the way up to $14,250.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 4:03 pm

fyre4ce wrote:
Tue Jun 11, 2019 3:45 pm
trueblueky wrote:
Tue Jun 11, 2019 3:28 pm
Real-world marginal rate > 100%
MFJ no children, AGI $38,000 reflects $2000 contribution to IRA. Tax before credit is $1403 per tax table. Couple receives savers credit of $1000. Tax = $403.
MFJ no children, AGI $38,001 reflects $2000 contribution to IRA. Tax before credit is still $1403. Couple receives savers credit of $400. Tax = $1003
One dollar increase in income yields $600 additional tax.

Real.world marginal rate <0%
MFJ 3 children, AGI $14,249. No tax due before applying credits. EITC = $6401
MFJ 3 children, AGI $14,250. No tax due before applying credits. EITC = $6431
One dollar more in income yields $30 more in refund.

If you include the end of the ACA subsidy at 400% of the poverty rate, one more dollar can mean $1000s more owed with tax return.
Thanks! Are there any examples of <0 or >100% marginal rates over a significant range, as opposed to a single spike?
For <0, there is the federal income tax rate due to the earned income credit (EIC) at very low earned income. See p. 23 of https://www.cbo.gov/system/files?file=2 ... rtbook.pdf. I believe the toolbox spreadsheet will show negative-changing-to-positive if cell B3 (salary income) is used for the x-axis and D63 (federal income tax) for the y-axis. If D67 (total tax) is used for the y-axis, FICA taxes counteract the EIC.

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 11, 2019 4:22 pm

FiveK wrote:
Tue Jun 11, 2019 3:33 pm
fyre4ce wrote:
Tue Jun 11, 2019 2:53 pm
How do you suggest we move forward?
Taking the draft versions as of this date and time and moving them to the public wiki would work.
The problem is that I, and the majority of other contributors in this thread, disagree with the way it's written now. I've tried to avoid an "edit wars" situation and to instead reach a consensus here through discussion. I'm worried we're at an impasse and am looking for a way to move it forward. I'm happy to consider reasonable compromises or other ways to resolve the disagreement. But, I don't think "just do it my way" is a good plan; you wouldn't appreciate that if I presented it to you.

In any case, the reason that most people in the personal finance community say the decision should be based on marginal rates is because for the simplest case, which helps people understand how the analysis gets done, that is true. But there are a lot of situations that can complicate it. When comparing investments of different performance/fees, it's no longer entirely about marginal rates, nor for cases when maxing out accounts. For the "irregular" (I'm open to a better term) rates analysis, I prefer to say that the average rate over a given interval should be used instead of the simple marginal rate, rather than redefine average to be marginal, so we can keep using the marginal term. I spent a lot of time explaining how marginal and average are actually two different things, and how using one term to describe the other is a mistake. I don't feel like those points have been adequately addressed. Even if we go with your proposal, we still need a term for (what I prefer to call) the marginal rate within an interval, if we're using marginal for the average rate over the whole interval. That was my question above, and we'd need an answer before we publish the page, to be consistent with terminology.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 4:48 pm

fyre4ce wrote:
Tue Jun 11, 2019 4:22 pm
FiveK wrote:
Tue Jun 11, 2019 3:33 pm
fyre4ce wrote:
Tue Jun 11, 2019 2:53 pm
How do you suggest we move forward?
Taking the draft versions as of this date and time and moving them to the public wiki would work.
The problem is that I, and the majority of other contributors in this thread, disagree with the way it's written now. I've tried to avoid an "edit wars" situation and to instead reach a consensus here through discussion. I'm worried we're at an impasse and am looking for a way to move it forward. I'm happy to consider reasonable compromises or other ways to resolve the disagreement. But, I don't think "just do it my way" is a good plan; you wouldn't appreciate that if I presented it to you.
Perhaps we can all agree that to count "majority," whether calculated for the way the article was modified 3 years ago, or today, might require using the fingers of both hands but probably toes are unnecessary. Including the references rkhusky added (just checked - they are still there), and including reasons while the textbook definition is good for the textbook but not real personal finance, seem examples of "reasonable compromise."
In any case, the reason that most people in the personal finance community say the decision should be based on marginal rates is because for the simplest case, which helps people understand how the analysis gets done, that is true. But there are a lot of situations that can complicate it. When comparing investments of different performance/fees, it's no longer entirely about marginal rates
For tIRA vs. Roth IRA, or t401k vs. Roth 401k, the investment choices will be identical so it is about marginal rates.
nor for cases when maxing out accounts.
It's still about marginal rates in that case. It's just that the breakeven withdrawal differs from the contribution rate.
For the "irregular" (I'm open to a better term) rates analysis, I prefer to say that the average rate over a given interval should be used instead of the simple marginal rate, rather than redefine average to be marginal, so we can keep using the marginal term.
Yes, we know what you prefer, but "just do it my way" is a two-edged sword.
I spent a lot of time explaining how marginal and average are actually two different things, and how using one term to describe the other is a mistake. I don't feel like those points have been adequately addressed.
I agree that "difference" is not the same as "derivative," but by defining marginal as a difference instead of a derivative seems to adequately address that point.
Even if we go with your proposal, we still need a term for (what I prefer to call) the marginal rate within an interval, if we're using marginal for the average rate over the whole interval. That was my question above, and we'd need an answer before we publish the page, to be consistent with terminology.
Again, it's not "my proposal" as in "I thought it up." It's a concept taken from another writer, suggested by a third person, that does seem applicable. I suggested "cumulative" as a compromise, because that is how the curve on the toolbox spreadsheet chart is labeled, but am certainly willing to consider others. I'd prefer not to use "average marginal" or "effective" due to the possibility of further promoting the "marginal vs. effective" misconception, but if we take great pains to distinguish those....

Topic Author
fyre4ce
Posts: 222
Joined: Sun Aug 06, 2017 11:29 am

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by fyre4ce » Tue Jun 11, 2019 5:23 pm

FiveK wrote:
Tue Jun 11, 2019 4:48 pm
I agree that "difference" is not the same as "derivative," but by defining marginal as a difference instead of a derivative seems to adequately address that point.
Even if we go with your proposal, we still need a term for (what I prefer to call) the marginal rate within an interval, if we're using marginal for the average rate over the whole interval. That was my question above, and we'd need an answer before we publish the page, to be consistent with terminology.
Again, it's not "my proposal" as in "I thought it up." It's a concept taken from another writer, suggested by a third person, that does seem applicable. I suggested "cumulative" as a compromise, because that is how the curve on the toolbox spreadsheet chart is labeled, but am certainly willing to consider others. I'd prefer not to use "average marginal" or "effective" due to the possibility of further promoting the "marginal vs. effective" misconception, but if we take great pains to distinguish those....
That I could get behind. It's consistent with the personal finance toolbox, and also avoids possible confusion with the more common use of "average" (average over total earnings). It also preserves the distinction between the rate for a small and large interval. :beer Apologies if I missed your earlier proposal. I have some other edits I'm planning on making; you OK with me doing a scrub of both articles and we can see how it looks after?

Oh, and an unrelated question - what is the capitalization convention for "traditional"? IRS Pub 590-A seems to use a lower case "t" although I've seen it both ways on the Wiki. I'll make it all lower-case, except for places where it would otherwise be capitalized, unless anyone has a better source.

User avatar
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 5:42 pm

fyre4ce wrote:
Tue Jun 11, 2019 5:23 pm
FiveK wrote:
Tue Jun 11, 2019 4:48 pm
I agree that "difference" is not the same as "derivative," but by defining marginal as a difference instead of a derivative seems to adequately address that point.
Even if we go with your proposal, we still need a term for (what I prefer to call) the marginal rate within an interval, if we're using marginal for the average rate over the whole interval. That was my question above, and we'd need an answer before we publish the page, to be consistent with terminology.
Again, it's not "my proposal" as in "I thought it up." It's a concept taken from another writer, suggested by a third person, that does seem applicable. I suggested "cumulative" as a compromise, because that is how the curve on the toolbox spreadsheet chart is labeled, but am certainly willing to consider others. I'd prefer not to use "average marginal" or "effective" due to the possibility of further promoting the "marginal vs. effective" misconception, but if we take great pains to distinguish those....
That I could get behind. It's consistent with the personal finance toolbox, and also avoids possible confusion with the more common use of "average" (average over total earnings). It also preserves the distinction between the rate for a small and large interval. :beer Apologies if I missed your earlier proposal. I have some other edits I'm planning on making; you OK with me doing a scrub of both articles and we can see how it looks after?
That's how wiki articles get built - have at it. :beer

See also When larger amounts may be beneficial, even though smaller amounts are not for one example of how to tell people when to do what. And I've always remembered something a writing teacher once said: "anything you write, I can improve; anything I write, you can improve."
Oh, and an unrelated question - what is the capitalization convention for "traditional"? IRS Pub 590-A seems to use a lower case "t" although I've seen it both ways on the Wiki. I'll make it all lower-case, except for places where it would otherwise be capitalized, unless anyone has a better source.
Lower case t. Roth is capitalized because it is a person's name.

User avatar
LadyGeek
Site Admin
Posts: 54913
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by LadyGeek » Tue Jun 11, 2019 6:14 pm

FiveK wrote:
Tue Jun 11, 2019 5:42 pm
Lower case t. Roth is capitalized because it is a person's name.
One minor correction. "Roth" is not capitalized because it is a person's name. See: Roth IRA
Named after US Senator William Roth, Roth IRAs were established by the Taxpayer Relief Act of 1997.
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
FiveK
Posts: 7005
Joined: Sun Mar 16, 2014 2:43 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by FiveK » Tue Jun 11, 2019 6:24 pm

LadyGeek wrote:
Tue Jun 11, 2019 6:14 pm
FiveK wrote:
Tue Jun 11, 2019 5:42 pm
Lower case t. Roth is capitalized because it is a person's name.
One minor correction. "Roth" is not capitalized because it is a person's name. See: Roth IRA
Named after US Senator William Roth, Roth IRAs were established by the Taxpayer Relief Act of 1997.
Ok, now I get it: it's not ROTH, but it is Roth.

rkhusky
Posts: 6753
Joined: Thu Aug 18, 2011 8:09 pm

Re: Seeking comment on proposed revisions to Marginal Tax Rate and Traditional v. Roth wiki articles

Post by rkhusky » Tue Jun 11, 2019 9:04 pm

It is worthwhile to note that these pages contain a lot of original research, e.g. all the tables and plots. If these pages were really Wikipedia-like, the vast majority of the text would be either a quote or a paraphrase of text from other publications. Perhaps the first section of the pages should be Wikipedia-like, i.e. heavily referenced, and a second section could include all the original research of Bogleheads' contributors. For example, a page the length of Traditional vs. Roth would probably have 30-50 references on Wikipedia.

Post Reply