I’ve seen cost estimates of the 401(k) deduction, something like $67 billion/year. Maybe it’s more, maybe less, I don’t know. My question is, does that cost only include the tax sheltered in a given year? Or, is that net of the taxes paid by people withdrawing their 401(k) and paying income taxes on those withdrawals?
Also, does it include the gain to the government by turning capital gains tax rates into income tax rates?
Further, I’ve heard that most of the benefit is going to upper middle-class or wealthy people. That’s probably true, in that many others can’t afford to save at all. However, lower income people who do save in a 401(k) will likely pay far lower taxes when they withdraw the money. However, a wealthy person will likely still have a high marginal tax rate when the money is withdrawn, paying back a large portion of the taxes deferred.
Finally, it’s my understanding that most people don’t really start saving big time in their 401(k) until they are older and have more room for such savings in their income. If that’s the case, however, than the benefits of tax-deferral (and the cost to the government) is smaller, since there are only a few years of deferral, and then the taxes will be paid at income tax rates rather than capital gains rates. Also, losses in a 401(k) aren’t deductible or subtracted from other gains like regular losses are.
Anyway, if anyone could shed light on this, I’d sure appreciate it.
A 401(k) or 403(b) plan can be a wonderful way for the average joe to accumulate a retirement nest egg, especially with a decent employer match. Even in the 15% bracket, the out of pocket cost is often a fraction of the initial investment amount. So while it indeed may benefit the wealthier more, we should be pretty careful when making any major systemic changes.
They seem to simply state the $100 billion cost to the US, since the paper is about retirement savings in Denmark.
I did a quick spreadsheet that tries to compare the costs and benefits of a regular brokerage account and a 401(k). I assume that the 401(k) investor invests $100, whereas the brokerage investor puts away $70 (the after tax amount, assuming a marginal tax rate of 30%).
I grow the 401(k) money by 6% for 10 years (to $179), then collect all the taxes, leaving $125 when the money is withdrawn. The government gets $54 in taxes.
I grow the brokerage money by 4.8%, which is the after-tax equivalent of 6% return, assuming a capital gains rate of 20%. At the end, the after-tax investor has $119, and the government has collected $40 in taxes (the initial $30, plus 1.2% of the growth each year.
So, the government gets MORE from the 401(k) investor than from the brokerage investor ($54 vs. $40). Sure, the money comes at the end, but in the steady state, with many people investing and retiring, that should wash out.
That analysis ignores the fact that capital gains taxes can be deferred as well. For example, if you invested in Apple stock in 1985, you still wouldn’t pay taxes on the growth, unless you sell it.
I guess I still don’t understand the argument that 401(k) is a net cost to the government, at least when it comes to high income retirees.
I think they key point is that most wealthy people plan for their retirement beyond their 401(k). That is, they invest pre-tax up to the point where it no longer benefits them and then invest post-tax to the extent that they can. The argument is that the 401(k) incentives are not doing much to change investment behavior other than provide a tax benefit to folks that are saving anyways.
In your example, the argument is that you wouldn’t see only $70 invested in the no-401(k) case but something closer to the full $100 because there is only so much money a person can spend - the rest is saved (either in a 401(k) or in some other savings instrument).
I believe most of the “cost” comes from the fact that most retirees have a lower marginal tax rate than when they were working and the time-value of money means that the government getting the cash now is worth more than getting it later.
Here’s a CBO report that puts the cost of “The exclusion of pension contributions and earnings” (which appears to include 401(k) and IRA benefits) at 1.1% of GDP. http://www.cbo.gov/publication/42919
The argument I’ve seen floating around is that the 401(k) benefits largely go to the wealthy. However, the wealthy would not likely see a drop in marginal tax rates during retirement, so they pay it all back eventually.
Regarding the time value, as long as you have many people saving and retiring at the same time, don’t the time value losses largely wash out? That is, sure, you miss out on the $100 of income from person A, but person B just cashed out $173 of 401(k) savings. Seems like the time value of the taxes would be a transient problem (assuming a stable population – even if the population were growing modestly, this would certainly be a second or third order effect, especially in a super low rate environment).
The CBO “net” number doesn’t seem to be net of taxes paid later or net of increased taxes due to all the income being taxed at income tax rates rather than capital gains rates. It appears to be net of the effects on payroll taxes (whether that’s positive for taxes or negative, I can’t tell).
I would argue that if you have retirement savings in excess of what you can put into an IRA or 401(k), you should have fixed income assets in the 401(k) and stocks in the excess, since stock investments already come with tax deferral and a lower tax rate.
But, back to the main question: Does the 401(k) and IRA cost to taxpayers really include the excess taxes collected later? Is there any way to find out what goes into the tax cost numbers?