gunny, most of the experts are predicting lower returns on account of
1) mean reversion in valuation (valuations are above average)
2) mean reversion in long term margins (margins are well above average)
3) recent low real earnings growth continuing
EDIT:
To provide a little more color
stock returns = dividend yield + real earnings growth + inflation + P/E multiple expansion / contraction
Most long term forecasts will make some assumptions about those 4 factors. Most things that I've seen that are predicting lower rates of real return assume
a) lower real earnings growth from margin contraction, mean reversion in margins, this is the basis for GMO's 7 year forecasts on stock returns being so low for U.S. stocks
b) mean reversion to historical valuations
If neither of those happen, then real stock returns from should be roughly equal to
2% (dividend yield) + 1.7% (LT real earnings growth) = 3.7%, right in line with your expectation
If valuations expand, returns will be higher. If valuations contract they'll be lower. If earnings grow faster than their historical real rate, then returns may be higher, etc.
The bottom line is that when an expert is making a long term return forecast, they are expressing some view of what research affiliates calls the "4 fundamental building blocks of stock returns" (dividend yield + real earnings growth + inflation + P/E multiple expansion / contraction)
Earnings growth can be broken down further into: revenue growth, EBITDA margin expansion / contraction, tax rate, leverage
P/E multiple expansion can be broken down further: bond yield & equity risk premium. You can make it as complex or as simple as you like.
https://www.researchaffiliates.com/documents/IWM_Jan_Feb_2012_Expected_Return.pdfhttp://fortune.com/2011/06/12/buffett-how-inflation-swindles-the-equity-investor-fortune-classics-1977/https://www.gmo.com/docs/default-source/public-commentary/gmo-quarterly-letter.pdf