The Voice over IP argument

A common argument about delivering Internet to the developing world, is that given IP networks, VoIP is practically free. Blog.org attributes it to my friend Howard Rheingold:

Howard Rheingold

The argument is very valid, at least in theory, but it has to be remembered, that while infrastructure for Internet access (IP) can be used at no additional cost for VoIP, there is a significant increase in difficulty/complexity.

I'm reminded of a meeting I went to at the Danish Unix-User group on the subject of VoIP. A technical run-trough of the Open Source solutions available for this purpose. The conclusion of this meeting of technically skilled unix users was that getting this to work in a reasonable manner, short of buying fairly expensive and proprietary IP phones and switchboards, was pretty difficult. It is still difficult to get a IP-to-copper and copper-to-IP link set up based on existing computer hardware and cheap handsets.

Now if you need to invest in dedicated VoIP equipment, that could still set you back over USD 500 for each unit (not counting the IP to copper gateway), which is  a lot of money.

Add to this, as David Brake (blog.org) correctly points out, that many governments are severely opposed to this application, and it gets still more complex. And remember that there is in fact, a good reason why many developing world governments are opposed. It's not just short-sightedness, corruption and unwillingness to change (although those certainly play a role as well). There is a basic difference in the way revenue from international traffic is distributed.
Telephony traffic is split (more-or-less) evenly between each of the telco's at either end, ie. quite a bit of revenue is generated for a national telco (or a commercial telco that then pays license-fees to the government). With IP networks, on the other hand, it is pure cost for the telco's, who are paying the full traffic fees to the company that owns the International IP lines.

Take as an example Ghana Telecom. Everytime an expat Ghanaian decides to call his family back home, a fraction of the fees he/she pays are given to Ghana Telecom (and hence to the governmet of Ghana), and the local Ghanaian has no expenses for that call. If that same relative decides to call using a VoIP systems, Ghana Telecom has to pay the price of the traffic to their International bandwidth provider, and hence not only lose the revenue from the phone call, but also have a Ghanaian pay money to a US or European company.

Now consider that in some countries over 15% of the Gross National Product is generated by the National Telcos it's no wonder that some governments get a wee bit protectionistic over that income.
(Thanks in large part to Andrew McLaughlin for explaining this situation to me).

Now, this should of course largely be balanced out by the fact that VoIP is an inherently superios (and cheaper) technology, and therefore would allow much better communications, especially those initiated by people in the developing world. But it is still important to consider this loss in any move to liberate the world of communications.

The ultimate conclusion of this is probably echoing  David Brake's, namely that this is a hugely complex matter, and we need more research!. But on a more practical level, we need to solve the complexities of setting up VoIP. There should be a simple to use, out-of-the-box, Open Source solution for VoIP, that way the political issues may still be there, but the high price, and complexity of getting this to work could be a thing of the past.