Ethiopia’s ETC: The elephant in the room slows down ICT development

Top Story

The Ethiopian Government was one of the first to embrace the use of ICT as a way to change Government and improve the efficiency of the economy. The country has a burgeoning ICT sector but it is being held back by the impact of Government policy. However laudable the Government’s intentions, there is an overwhelming mismatch between its rhetoric and the results. Our correspondent takes a look at the elephant in the room that isn’t being dealt with.

There is an apocryphal anecdote about an Ethiopian policy-maker who was asked when his country was going to liberalise. He replied:”We’re waiting to see what happens first before we decide whether to go ahead.” Ethiopia is very different from the rest of Africa.

Like the Soviet Union, Ethiopia went from feudal monarchy to Marxist revolution without going through capitalism. This history casts a long shadow even today, particularly in terms of how the Government sees issues. The Government’s bureaucratic processes are painfully slow. Nevertheless, the country runs one of the most successful airlines on the continent and has a cadre of extremely bright and well-educated people.

So it is not altogether surprising that the green shoots of the free market are beginning to show through in the ICT sector. There are a clutch of small-scale entrepreneurs doing software and networks and a couple of outsourcing companies that you would not believe could have started here. These companies are run by a number of home-grown entrepreneurs and some diaspora Ethiopians.

One of Apposit’s partners – Adam Abate – is one of those diaspora Ethiopians who has returned. His company has developed online financial software and his client list includes Addis Ababa University’s Data Centre, the Ministry of Finance, Ethiopian Airlines and several banks.

As he sees it: ”There are lots of business opportunities here. The IT sector is under-served and serviced very poorly. There is not enough scale to get the breadth of experience required. (Nevertheless) the Government is very supportive of ICT and wants to move 100% towards it.” He notes that IT salaries have increased in the last two years.

Geomark is an Ethiopian-run company that takes information off satellite maps and turns them into usable information for digital maps and GIS applications for a European company: we will return to look at it in more detail in a later issue. It started with a small project which it did to demonstrate it could be done and is now a significant supplier. With all the disadvantages Ethiopia has, it still reckons that it is 2-4 times cheaper than equivalent suppliers in China and India. There is also another company called Eventive doing outsourcing work.

Solomon Tesfaye’s e-Systems Africa provides portals, web-based applications and training and next month will become an emc2 partner. Its clients are mainly Government, the Ethiopian ICT Agency and NGOs (Care Ethiopa).

The Ethiopian ICT Agency was set up to encourage the process of ICT implementation and is focused on getting Government ministries to do it. It has been instructed by Government to prioritise the Education, Health and Agriculture Ministries. It is also rolling out telecentres to rural areas and is building an ICT park in the capital. Located at one end of the international airport’s runway, it has been allocated B20 million this year for the project and has started putting in the roads, sewers and service buildings. It has been struggling to get the project done but reckons it will be complete in about a year and half. Companies locating there will be promised dedicated bandwidth.

All of the above can be counted as success….However, you know there’s a but coming in the sentence. All of this seeming success is being held back by the poor services and high prices of the monopoly telecoms provider ETC.

As one person told us about the quality:”It’s horrible…The prices charged are extortionate. The internal routing within ETC is horrible which is the reason why the Internet is so slow.” Another told us:”The connection speed is not good. Web designs need to be light weight to work.” Because of speed problems, Geomark is unable to get small-scale, fast turnaround work because all their existing larger-scale work is sent to them by DHL. Clearly the ICT Park with its promised dedicated connection is an attempt to deal with the barriers that ETC has created.

The Government’s much trumpeted Woredanet – a Government network run by the Ethiopian ICT Agency using fibre and satellite which connect 600 Woredas (provinces) – suffers a number of problems when used for anything but video-conferencing. Someone close to it told us:”It’s got bandwidth problems when running.” As a result, the Government is expanding capacity and increasing the number of nodes by 100. It may also start supplying IP telephony.

Private sector companies only wish they had some of this kind of national networking. Private banks have expanded a great deal and started putting in ATMs in the capital and branches outside of the capital Addis Ababa. The bandwidth they have is expensively priced and service is awful. When a connection to a branch outside the capital goes down it can take a long time to fix and meanwhile transactions have all to be laboriously verified by phone. As one banking person told us:”As a result of the bandwidth problems, there’s no hosted applications.”

At an international level, although ETC has a fibre connection via Sudan that was opened in 2007, the country seems to have extremely slow Internet connections. As an admission that this is an issue, 20 exceptional licences for VSAT connections have been issued, one of which was secured by Ethiopian Airlines which argued that it was essential for the success of its business. ETC has 1,700 corporate customers.

The poor old Ethiopian consumer is also not doing much better. ETC’s DSL customers (1341) and CDMA 2000 data users suffer daily outages and customers often keep a dial-up line as a back-up. Some even say that the dial-up service is better than any of the newer services. A 128K DSL line costs an eye-watering US$360 per month.

ETC charges B11 a minute to Washington DC whereas similar calls in grey market outlets cost B2-4 a minute. Periodically the Government raids premises, seizes equipment and closes places down but the grey market does not go away. A study commissioned by ETC in 2002 recommended that international rates come down to B5.5 a minute.

Mobile customers (1,984,000 in June 2008) are faring little better. Because of investment constraints and poor equipment and billing decisions, capacity has not kept pace with basic voice demand. According to one insider:”When it started, ETC did not do any serious frequency planning. It put up base stations on a one-by-one basis. There are still coverage black holes in Addis Ababa.” Nevertheless ETC is now the proud owner of a 3G network in the capital but announcing cutting edge technologies whilst not actually dealing effectively with current demand is a repeating pattern.

So how did this all happen? ETC was one of first adopters of fibre networks and its implementation became a Cisco case study. It has a fibre network of 4,000 kms and plans to build a total fibre network of 6-10,000 kms in two years time. The fibre radiates out in four directions from the capital. It has a fibre ring in Addis Ababa and in some provincial cities and has started the transition to soft switches. It was ahead of the pack in East Africa when it opened an international fibre link via Sudatel in neighbouring Sudan in March 2007.

However it offers some of worst quality and highest priced bandwidth in Africa and believe me there are many awful places it had to be worse than to get into that league. Also the cost of mobile calling is at the high end of the range.

Whilst sources in ETC claim that the link via Sudan has considerable capacity, those who should know say that it is a maximum of 2 mbps and the only alternative is a 33 mbps microwave link via Djibouti. But another well-informed person told us:”Most of the voice and data goes out via satellite. Either capacity on the Sudan link is low or its unreliable.”

The link to Djibouti was supposed to have been in place 12 months ago. However, ETC proposed to that other great regional monopoly Djibouti Telecom that it would co-locate at the landing station. French advisors to Djibouti Telecom thought this was a bad idea and it became a Government-to-Government political issue that has not yet been resolved. SEACOM suggested a link via Somaliland but that has proved politically difficult to achieve. However, both ETC and Telkom Kenya are near the border which would allow access to TEAMS and SEACOM but again there does not seem to have been sufficient focus on solving this problem.

But there is more fundamental problem needing to be addressed. ETC is a state-run telco that whilst the Government says that it is profitable, it does not render accounts that include details like capital depreciation. So leaving aside management and staffing issues, ETC has been starved of capital, forcing it to rely on an often incompatible mix of vendor-financed equipment. If the sector was liberalised (as it has been in neighbouring Kenya and Sudan), it would have attracted something like US$1-2 billion in investment with all the nice things that occur when you spend this kind of money in a country.

In the meantime it has the worst of both worlds in terms of tariffs. For as one insider who knows ETC processes well told us:”It doesn’t put a lower price on broadband because it would drive up demand too fast.” In other words, it is rationing shortages by price whilst all the while Government is talking about a policy of implementing ICT at all levels.

On the other side, employees of ETC still get free services from their employer. But as one private sector person wryly observed:”They should (pay and) share the pain of their customers.” Even Woredanet does not pay for its bandwidth: it simply tells ETC of its requirements.

The Government’s answer to this (for it is more or less indistinguishable from ETC at the level of strategic decisions) has been to take US$1.5 billion loan from Chinese vendor ZTE to use for network expansion over four years. There are two possibilities that will result from this decision. Either ETC will have to behave in a far more commercial fashion if it is to meet this level of repayments or it will default and the loan will be rolled over.

So last but not least this brings us to the thorny topic of whether ETC will be privatised. World Bank pressure and funding in the broadcasting area seems to have persuaded the Government to allow some modest opening of FM stations and local community radio stations. But with ETC you get the sense that they don’t want to let go if only they could find a way of avoiding the inevitable.

ETC is now working with private companies selling airtime but it put the idea of independent ISPs on the table (licences were even issued) but withdrew the idea. Periodically someone tells us that they will be opening a mobile operation in Ethiopia but nothing ever happens. Whilst in other countries of a similar size there are many different sized companies spreading skills and ideas, in Ethiopia there is only one company managed by the Government.

Apparently the Government’s formal position is that it will be privatised after the rural areas are fully connected. Whilst this may sound like a “this year, next year, sometime, never” answer, sources say that rural connectivity could be completed within a year. Others talk of a timetable of 1-2 years but the problem in Ethiopia is that you have to remain optimistic. This is not a decision that is publicly debated but one that will be taken by one man: the Prime Minister. The next election is just under two years away and therefore the decision probably won’t be made until after that date.

For as one private sector person told us:”People in Government know the telecoms monopoly is blocking the development of the ICT sector and whole economy. The market may persuade them to change.”

Note: At the time of writing, 1 Ethiopian Birr = US9.8 cents.