It seems to me that Roth contributions to a 401(k) or a Roth IRA could be extremely tax-advantaged over pre-tax 401(k) or traditional IRA contributions, but at the same time, I have trouble believing that analysis.
Here is my analysis:
Assumptions: Let's assume that the marginal income tax brackets and rates never change in the future. And let's assume that my income never changes either, at least before I retire. Today, my annual income is such that my overall tax rate is 25%.
Scenario 1: I contribute $1,000 pre-tax to a 401(k). Assuming it grows at 5% annually, after 25 years I now have about 3.386 times what I originally had, or $3,386. I then retire and start withdrawing the money just fast enough to match my pre-retirement annual income, until it is all gone.
By my understanding, pre-tax contributions to a 401(k) or IRA and the earnings on them are taxed on withdrawal at the ordinary income tax rates. So in this scenario, I have $3,386 in taxable income from my 401(k).
Since I withdrew my 401(k) money at the same rate as my pre-retirement income, the overall tax rate I face while I withdraw that money is still 25%. So I pay about $847 in taxes.
Scenario 2: I make a $1,000 Roth contribution to a 401(k). Since I can't deduct Roth contributions from my taxable income, I separately pay $250 tax on the income I contributed.
The money grows at the same rate as in Scenario 1, and I also retire after 25 years with $3,386 in the account. And like in Scenario 1, I withdraw the money just fast enough to match my pre-retirement income, until it is all gone.
Earnings on Roth contributions are not taxed ever. So, my $1,000 Roth contribution is taxable income on which I've already paid tax, but the $2,386 in earnings over 25 years is not taxable income. This means my total tax burden is $250.
Conclusion: In Scenario 1, my total tax burden is $847, while in Scenario 2, it is $250. That is $597 less, or over 70% less tax.
So based on this back-of-the-envelope analysis, it seems that I would save a lot of money in tax payments by favoring Roth contributions over pre-tax contributions. And yet it feels like Roth contributions would be orders of magnitude more popular than they are if that were really true for most people. Hence the question:
Have I gotten the math wrong in my analysis, or are there other things I am missing (such as in my assumptions)?