Why are you and others continuing to push the "50% of *gross*" false angle? MMM uses net income in his "shockingly simple math" calculations, not gross. Anyone saving 50% of gross is going to hit FI a lot sooner than 17 years, let alone find some ugly surprise down the line as you've suggested.

The only reason we're talking about percentage of gross is because the OP asked about it that way. We should have cleared up the misunderstanding more clearly before this point. Thanks for finally doing that.

Velocistar, is that in the ballpark of what you actually pay after credits and deductions?

Good question. I have access to my paycheck right now but not my tax return. We got decent refunds last year, so I set my exemptions sky-high in February or March, after the withholding was already high for a while. So, we pay a bit less than what I just posted, especially in federal, but the numbers are in the ballpark. Maybe 1.5% for federal, making it 10% overall. The state number should be pretty accurate, and obviously the FICA number, too.

As for the topic of whether savings rate is a reliable predictor for time to retirement, many of the things we've discussed are things like increasing expenses, changes in tax law, etc., which are just changes in the expense side of the savings rate. That's not the model itself, it's just an input to the model. Once you make that adjustment, the model still holds as a useful prediction given your best information. No one is saying that if your initial savings rate is 50%, then your time to retirement is 17 years no matter what happens after that. As others have pointed out, this can work in your favor as well, since you do have a large degree of control over your expenses.

There's one element of the model that I think is extremely reliable, and that is the effect we're discussing about changes in savings rate. For a person starting out with nothing, with a 4% real return, going from a 10% savings rate to 20% will bring retirement about 17 years sooner, from 20% to 30% about 10 years sooner, 30% to 40% about 7.5 years, 40% to 50% 5.7 years, and 50% to 60% 4.7 years. That's about 46 years total going from 10% to 60%. These are differential effects that would propagate, within a factor of 2 or so, through any set of market returns. The differential effect is greater the worse market conditions are, so the advantage of higher savings rate over lower savings rate is even greater when returns are low. It's quite a powerful model, both financially and mentally.

For those who are not just starting out, we've heard several stories of people who come to realize that after they reduced their expenses, they could already retire. Brave New Life comes to mind.

You know, this whole discussion is just a variation of the unending 4% vs 3% SWR argument, except it's about the accumulation phase. If you're really concerned about it, be conservative and work the 2 extra years to go from 4% to 3%. FIREcalc shows us that a 3% SWR is quite conservative. SWR is actually an input to this time-to-retirement model, just one that MMM freezes at 4% in his blog post for simplicity (and because he believes that it's already conservative). Choose different values if you want, here's the model:

T = ln(1+ROI*(1-SR)/(SR*SWR))/ln(1+ROI)

T = time to retirement

ROI = return on investment during accumulation, in real terms

SR = savings rate

SWR = safe withdrawal rate

To sum up the effect of an uncertain future on time to retirement vs. savings rate, I have only one other thing to say.

YMMV.