I think the retail people like you and us make up less than 5% of the total hashrate. Those higher up in the foodchain probably get equipment more than 50% cheaper than us so they can continue to mine for awhile and the rate still goes up because there is competition between the suppliers who can make more and faster, wafers, chips, boards, kits, etc. As the margins get squeezed for even the manufacturers, those that are the slowest to develop newer and faster and cost efficient equipment will falter.............. BFL, Black Arrow and KNC soon.
Which brings the topic of the operational costs. If a vendor gets a miner at $0, then they still need to run it. They will pay electricity (20% cheaper than residential costs on average), cooling, personnel, rent, taxes, etc. From those operational costs, about half is electricity.
If network hashing power goes so high that manufacturers can not run their miners at a profit (eg. Selling the mined bitcoins pays only 2x the electricity costs, which barely cover its costs) then even manufacturers will stop at that point to put more miners in the network. If there is no profit on it, there is no investors money. There is a "cost of opportunity" of money. No investor puts money in a high risk enterprise without expectation of a big return.
Also consider that half the current hashing power is made up of old tech ASICS, which consume 2W/GH.
The other half (1W/GH - 1.2W/GH) started shipping only late 2013 in limited quantities, as there are not many factories worldwide that can produce massive amount of 28nm asics.
The new generation 20nm ASICS (
KNC Neptunes, review here), will consume 30% less than 28nm ASICS (according to KNCMiner). That means about 0.7 W/GH. Then 100% of the current network hashing will be obsoleted long before your Neptune is out of business.
28nm ASICS (whichever vendor) are also top of the line, and only bested by the new generation 20nm Neptunes.
Then retail miners like us, if using 20nm-28nm asics, still have a chance against vendors mining themselves. Not like in 2013, but still barely profitable to have it as a second steady income besides our daily jobs.