Now you see why Satoshi used a subsidy to bootstrap the economy. You can also see why the subsidy can decline. Remember Bitcoin is a work in progress (BETA) so nobody has all the answers.
Today if there was no subsidy for miners to make the same (in USD terms) would require a fee of roughly $2.67 per tx. Ouch obviously not viable (still cheaper than WU though ). The good news is that roughly a year ago the cost per tx would have been ~$6.00 per tx and roughly 3 years ago the cost would have been closer to $50 per tx.
http://blockchain.info/charts/cost-per-transaction?timespan=1year&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=
If tx volume continues to increase then the unsubsidized cost of the network gets cheaper per tx.
VISA processed roughly 80 billion tx last year. If Bitcoin was someday 10% of the size of VISA that would be ~8 billion tx annually. At an average tx fee of $0.01, it would be the cheapest global payment network and would produce ~$80 per year in revenue for miners. Today with a subsidy of 50 BTC miners are compensated ~$28 mil annually at current exchange rates. Will Bitcoin grow that large? I don't know just pointing out a small fee on a lot of tx can be a very large number.
This is exactly what i think needs to happen for bitcoin to survive the next 50 or so years, maybe even the next 10 . A lot can happen in 5 decades. Im just unsure how a decentralized currency can work along side centralized ones (just sayin again i have limited economics knowledge limited to micro/macro-econ college classes :p) Bitcoin tx would have to grow by like 300% to reach a tenth of visa so idk, it would probably need a BIG niche to fill in.
I hope for the best though for btc cause i love it and want to see it built by real pros.