[kictanet] Day 3 of 10:-IGF Discussions, Internet Interconnection Charges

[Michuki Mwangi]

David Otwoma wrote:
> Alex,
> 
> One has to agree with your argument.
> 

> Remember when mobiles were costing above 200k shs a set. Then the
> regulatory environment allowed more than one player and a plethora of
> sets. We now have 3 shs a minute calls and sets not above 1000 shs.
> When more players roll out the 11 pm to 6 am free calls will go to
> free calls on weekends and public holidays.
>

Caution here .. when you talk of mobile calls - you are referring to an 
operator who owns a network and charges you for minutes you terminate 
within their own network. Imagine if you were browsing with your ISP and 
didnt have to exit their network and they owned their own-infrastructure 
or links to your premise - exactly how much do you think it would cost?. 
The costs will always scale upwards the moment you dial a number outside 
the mobile operators network.

So this is a perfect example of what interconnection (read peering) 
should do to costs of Internet services. If the ISPs would be able to 
provide a local Internet service only - am certainly sure it would cost 
less than 1,000 Kshs a month for unlimited broadband links. But then 
again what would you use it for?


> As for ISP, when even GSM (Zain has a 2,995 shs unlimited service
> while Safaricom has 300 MB bundle for 999 shs) have joined the earlier
> players (dial as in Telecom, Popote etc. wireless as in butterfly,
> iBurst etc. and the conventionals), the trend has been lower costs
> with better quality as the consumer has many choices.
> 
Maybe i need to rephrase myself on this one. Competition is good for the 
customer because he gets a variety of options. The options allow him to 
select a provider that suits his pocket. However, that comes at a cost 
to the user with regards to the best effort service that they receive. 
In most cases the contention ratio will be higher as the prices goes 
lower. In recent case of Comcast, i would understand why they would 
filter out access to torrent/peer-to-peer applications. This is because 
about 5% of the users would consume over 50% of the bandwidth available 
therefore impacting negatively on the calculated contention ratio for 
the entire network. In a competitive environment that is driven by Costs 
vs. Qos you want to maintain your service quality for the 95% of the 
clients on the network and cap out those who are seen to "abuse" the 
network. After all bit-torrent is known to provide access to alot of 
"illegal" content.

Also have you ever wondered why the service is impressive during the 
first 6 - 12 months of introduction of service and he quality sort of 
degrades with time to a point that its no longer worth considering 
despite the lower cost. Then the most common thing happens, a new 
provider or an existing provider launches a new product which is often 
lower priced compared to anything thats in existence in the market and 
you have this migration of folks for the next 6 - 12 months. After that 
the service quality goes down and the same pattern keeps repeating itself.

That to me is not a good thing to the consumer for the following reasons;

1) It means the user has to change provider every year which is a huge 
inconvenience

2) The cost of setup in most cases means that the user builds a store of 
CPE that they may never resell or make use of - i have one modem with me 
that i cannot sell, am sure there are those with more than two.

3) 2 above means that the cost of investment per modem for the user per 
year is quite high and hence does not actual calculate into lower costs 
to the user but high costs if thats anything to go by.

4) Where there were contracts to be signed - terminating the contract 
can often lead into additional costs for the end user.


> Sectors that support ICT like energy are the let downs as they are yet
> to liberalize i.e. Kengen has no serious competitors and KPLC is still
> a monopoly giving many Kenyans a raw deal. Those escalating energy
> costs are reducing the impacts that would have been felt by growth in
> ICT.
>

I agree with this and not only ICT is suffering a good number of sectors 
have been affected equally. But is it only because of lack of 
competition? I tend to think that we lack the motivation and interest to 
to go beyond the textbooks. I think the lack of creativity and 
innovation in most of our developing countries is a significant 
contributor to the status we have in ICTs.
For instance in China where the Internet is highly censored the number 
of users has reached 253million surpassing the US. The following article 
  might be a useful highlight to what innovation can do in such highly 
regulated and restricted environments.

http://www.iht.com/articles/2008/07/25/business/internet.php

Regards,

Michuki.