Obviously the lump sum would probably be able to yield that sort of money in interest, but the interesting thing is that it should be much easier to secure the monthly passive income than it would the lump sum.
And the higher the payout the less passive/safe it becomes (i.e., CD > Real state > businesses > web apps > ???)
And "time" is work. Building a house also requires just some materials and some time :). Even real state needs management (things break, rents have to be collected, risk of lawsuits, tenants that dont pay).
That said, $5,000 a month is $60K a year, or 6% of $1,000,000. I'd probably take the $5k also since it's a reasonable (and 100% guaranteed) rate of return on the lump sum. I wouldn't feel much safer because it's "cast iron" -- it could still be worthless in 30 years. It's just a slightly better deal in your example.
The long term 7% average often quoted isn't if you average the last 15 years rather than the last 150.
From a quick check of Yahoo Finance, it looks like the Dow dropped from 14k in July 2007 to 8k in January 2009. Sure, a fund is meant to outperform the market as a whole, and sure, most of them don't manage that for obvious reasons, but when the market plunges like that no-one's going to be safe.
No, that's not being fair at all. That's being irresponsibly optimistic. Let's say it is now 2005, you have a $1,000,000. You follow cpursley's advice, and put it in a well managed ETF/mutual fund, expecting to get 8%-12%. You get that for the first year and second year - and then you get -50% (which, if you look at your starting sum is "just" -40%). Now, you need 8 good years just to go back to where you started.
2005 is unfair? Go back to 1999, and check yourself at the end of 2001.
1999 is unfair? Go back to 1986.
My point being, the "common knowledge" 150 years long term average of 7%/year is irrelevant, for two reasons:
a) market dynamics have changed significantly in the last 10 years or so; the 140 years before that aren't as informative as the last 10 - and the last 10 are abysmal.
b) even if the 150 year statistics were relevant - what you care about is your 10-30 year horizon. There are just too many unlucky events during those 150 years that you are very likely to have a 30-50% loss in one of the years of your 10-30 year horizon -- which might take you 50 years (that you don't have) to average out and come out on top.
Anyone who really believes the 8%-12% numbers can borrow at <6%, invest at >8%, leverage ad absurdum and basically print money. Which, incidentally, is the empirical proof that 8% is not attainable.
(P.S - CALPERS assumed a modest 8% return. That's why they're now only 50% or so funded http://www.calwatchdog.com/2012/03/22/calpers-funding-might-... . If you can get them 8%, they'll give you $20M/year in salary. Really.)