Certainly the "survivorship bias" inherent within the index has an effect. So the losers get booted out, the winner stay in and grow. But since the market is usually defined as "cap weighted", the higher cap stocks have a much bigger impact on the value of the market than some microcap stocks that are worth next to nothing.
As to whether this can be tested or proven? Easily. If you had the raw data listing all stocks (including the microcaps), you could simply add up the market cap of all stocks 20 years ago, then do the same today, and show that the market grows over time.
Figures about how much the market grows overtime really do depend on when you set your start and end dates, and whether you include dividends. Let me pull out the raw figures and do the math for you.
So now we can confidently say that the All Ordinaries, including dividends, has returned 11% over its lifetime, and we have the math to prove it.
