> Sure, but after people stop working it's common for them to replace their working income with Social Security, other pensions, and withdrawals from retirement accounts in which their savings compounded over time.

Agreed, but there was no room for savings in your model as you used the whole income as spending (median income * 60 years). Taxes makes a big difference too since it doesn't count as spending. E.g. someone that earns 55K$, pays 25K$ in taxes and saves 10K$, only spends 20K$ per year, amounting to 1.2M$ (20K$ * 60 years).

$25k in taxes on a $55k income? Where are you from?

For a family in 2024, at the Federal level, the first $29,200 is untaxed because of the standard deduction. The remainder would be taxed at 10-12%. Child tax credits can easily wipe that out and make the liability negative (i.e. you get all your tax withholdings refunded plus another couple thousand dollars).

State income taxes will vary, but are typically lower than Federal income taxes. Many have no income tax.

Just picked an arbitrary number to illustrate. Regardless, we are far from the $3.3M figure. The point is simply that median income * 60 years is not a reasonable approximation for lifetime spending.

But they're also counting retirement as an expense category. Which would imply that money is being spent twice?

Good catch; I overlooked that line. I think we can take that as further proof that this article is not worth our attention.

"401k" is obviously not an "expense" independent of the things that 401k savings are actually spent on (I mean, unless one is counting the expense ratio on their mutual funds).