> This is analogous to claiming that microwave ovens increase the risk of cancer, even though we can't detect it.

To be fair, the control group would be difficult to say the least. The same is true with the economy. So yea, it's very possible there are behaviors we can't detect.