Home / mbt worldwide bus / 0 - PPIAF

0 - PPIAF - mbt worldwide bus

Public Private Infrastructure Advisory Facility

Study of urban public transport conditions in

Addis Ababa, Ethiopia

IBIS Transport Consultants Ltd

March 2005

0 Introduction 1

0.1 Context of the report 1

0.2 Currency of reporting 1

0.3 Ethiopian calendar 1

1 Political context 2

1.1 Institutional structure including allocation of powers between jurisdictions 2

1.2 Central government policy position on private supply in public transport 2

1.3 Existence of a private sector in other public utility sectors 3

1.4 Local government policy stance and willingness to reform 3

2 Economic conditions 4

2.1 National GDP per capita 4

2.2 Percentage of income spent on public transport 4

2.3 Current transport fares and costs, in current dollar equivalents 4

2.4 Population of Addis Ababa 5

3 Current public transport patronage 6

3.1 Passenger numbers by category 6

3.2 Ticket types 7

3.3 Passenger load factors 7

4 Structure and organisation of urban public transport 9

4.1 Organisational structure 9

4.2 Legal basis for organisation and jurisdiction 9

4.3 Vehicle ownership and management 10

4.4 Vehicle purchase, finance and insurance 10

4.5 Legislative instruments 11

4.6 Staffing and skills analysis 11

4.7 Facilities and equipment 12

4.8 Financial environment 13

4.9 Role of the informal sector 14

5 Service supply characteristics 16

5.1 Operational characteristics – bus services 16

5.2 Operational characteristics – minibus services 18

5.3 Current bus routing and scheduling plan 21

5.4 Income and expenditure statements 22

6 Vehicle related data 25

6.1 Vehicle ownership 25

6.2 Constraints on supply of sector inputs 25

6.3 Financial arrangements for fleet procurement 26

6.4 Fleet inventory 26

6.5 Fleet maintenance 27

6.6 Vehicle operating costs 28

7 Existing regulatory arrangements and institutions 29

7.1 Review of market conditions 29

7.2 Fare and fare change mechanisms 30

7.3 Effects of unions and union regulations 31

7.4 Taxation and other incentives / constraints 31

8 Perceived problems 33

8.1 Inadequate service quantity 33

8.2 Low levels of safety 34

8.3 Poor service quality 34

8.4 Low level of affordability 36

8.5 Low operating speeds 36

8.6 Poor service accessibility 37

8.7 Inappropriate vehicle type and size 38

8.8 Poor vehicle condition 39

8.9 Inefficient operating procedures 41

8.10 Inefficient network design 42

9 Attributed causes 44

9.1 Poor operating standards and inadequate enforcement 44

9.2 Lack of profitability and investment 45

9.3 Low skills base 45

9.4 Violent or illegal behaviour in the operating industry, and administrative corruption 46

9.5 Lack of economic regulation, network planning, and supporting institutions 46

9.6 Inappropriate ownership structures or company size 47

9.7 Inadequate transport infrastructure 48

9.8 Lack of empowerment of transport users 48

10 Reform programme 49

10.1 Nature of planned reforms 49

10.2 Drivers of the proposed reforms 49

10.3 Obstacles to reform 50

10.4 Linkage to other development initiatives 50

Appendix 1 51

Anbessa Route Analysis – January 2005 51

Appendix 2 56

Anbessa Financial Statements – Management Accounts 56


Context of the report

This report on the current situation with regard to urban passenger transport within Addis Ababa has been prepared as a reference database to assist with the construction of a ‘toolkit’ on bus transport reform to be financed by PPIAF under the direction of the World Bank. This toolkit is intended to serve as a manual for governments when they recommend and implement processes to reform and re-organise the provision of urban transport services.

The report seeks to describe the current arrangements for the provision of public transport in Addis Ababa, and analyse the deficiencies observed in the supply of these. It is based on two recent studies (one completed in the past year, and the second ongoing), together with further interviews of key stakeholders undertaken in the light of these.

The first of these studies titled ‘Improving urban transport through private participation in Addis Ababa’ was undertaken by LEA International of Canada and completed in February 2005. This study focused on an evaluation of the policy, regulatory and institutional options for improving bus services in Addis Ababa. A new regulatory regime for the public transport sector is proposed, together with measures to revitalise the incumbent publicly owned large bus operator.

The second study covers urban transport planning and traffic management for Addis Ababa, and is being undertaken by Consulting Engineering Services of India. Their initial Findings Report is due later in March 2005, but some preliminary material has been made available for this report. The study contains a detailed analysis of road traffic in the city, as well as of both the public and private sector providers of urban transport services.

Currency of reporting

All monetary figures used in this report are expressed in Ethiopian Birr. At the time of writing in March 2005, the rate of exchange of the Birr to the US Dollar was 8.657 to 1. This rate reflected an effective Dollar devaluation of less than 5% over the past year, enabling direct comparisons between each of the two studies referenced above.

Ethiopian calendar

The Ethiopian calendar varies from standard international practice in a number of ways. The Ethiopian year starts on 11 September in the Gregorian calendar, and the current year is 1997. There are 11 months each of 30 days, followed by one month of 35 or 36 days depending on whether there is a Leap Year in the Gregorian calendar. To complicate matters further, the fiscal year in Ethiopia starts on 8 July in the Gregorian calendar. However the fiscal year is given the number of the calendar year that it largely overlays. Thus the current year running from 8 July 2004 to 7 July 2005 is referred to as Budget Year 1997 EC.

Where appropriate, Ethiopian dating has been converted to the Gregorian calendar for the purpose of this report. Also, Ethiopian months have been converted to their nearest Gregorian equivalent in the text.

Political context

Institutional structure including allocation of powers between jurisdictions

The present institutional structure of the Federal Democratic Republic of Ethiopia is of relatively short provenance. The country had been ruled by an autocratic monarchy until 1974 when this was overthrown in a military coup. The ensuing regime, known as the Dergue, became increasingly dictatorial and repressive and was in its turn overthrown in 1991. A Transitional Government was then formed that laid the basis for the current constitutional arrangements. These were ratified by a Constitutional Assembly in late 1994, and took effect in 1995.

The federal arrangement, established under the Constitution, has guaranteed the rights of the Regional States to administer their own affairs. They are empowered to formulate policies that are appropriate for their respective development; to lay the foundation for economic and social infrastructures; to participate directly in sectors that are critical for their economic development; and to safeguard law and order in their own areas.

The Federal government and the Regional States have legislative, executive and judicial powers. The Federal republic has two houses – the House of Peoples’ Representatives and the House of the Federation – and nine Regional States who have powers to formulate policies and administer their respective states. The House of the Peoples Representatives is the highest authority in the Federal Government while the Regional State Councils have powers to legislate and act upon issues that are under their jurisdictions.

The president of FDRE is the head of the republic and is elected by a joint session of both houses for a term of six years. The highest executive powers of the federal government are vested in the Prime Minister and the Council of the Ministers who are accountable to the House of Peoples’ Representatives. This has the mandate to fix the term of the Prime Minister and to approve the ministers who are nominated by the Prime Minister.

The Constitution requires establishment of an independent judiciary, and supreme federal judicial authority is vested on the Federal Supreme Court. The judicial powers for the federal and state levels are vested in the courts.

Elections are held in the country through the guidance and execution of a National Election Board that is accountable to the House of People’s Representatives. As of 2004, there are over 60 political parties registered at the national and state level.

The Federal Democratic Republic of Ethiopia (FDRE) comprises the Federal government and the governments of the Regional States. The republic comprises of nine regional states that were established on the basis of settlement patterns, language, identity and the consent of the people concerned. In addition there are three city-states, of which Addis Ababa is by far the most significant. The City of Addis Ababa is wholly surrounded by the State of Oromia, but forms no part of this.

Addis Ababa is divided into 10 sub-cities for administrative purposes, and these are each further subdivided into kebele (‘wards’) as the lowest administrative tier; there are currently 99 kebele following some recent consolidation. The Transport Authority plans to use the sub-city as its basic unit for organisation of the public transport sector.

Central government policy position on private supply in public transport

The governing legislation for urban passenger transport is the 1992 Proclamation to Provide for the Regulation of Road Transport, brought in by the Transitional Government after the overthrow of the Dergue. Subsequent legislation allows each Regional administration (including the City of Addis Ababa) to develop its own regulatory framework, but this opportunity has not yet been availed.

This legislation effectively deregulated the supply of urban transport, but its implementation has so far been limited to the ‘taxi’ sector comprising of vehicles carrying up to 12 persons including the driver. Anbessa City Bus Services Enterprise, as part of the Passenger Transport Corporation, had held an exclusive franchise for urban bus services within the country. This has not yet been challenged under the proclamation, but it is understood that certain potential investors in the sector had been discouraged from involvement in the past.

Following recommendations from the study on improving transport, the Transport Authority of the City administration has decided to implement a limited competition regime for delivery of urban bus services but retain a self-regulated private taxi sector. At this stage it is not clear just how such a regime might be implemented, and whether the public capital and operating support provided to Anbessa will prevent the emergence of genuine competition. Either an additional market tier will need to be created (premium buses), or else the same level of public support will need to be made available to all potential market entrants.

There is also a lack of clarity regarding the ownership of Anbessa under the current federal arrangements with their emphasis on decentralisation of powers and responsibilities. Whilst Anbessa is a federal enterprise, its only operation outside of Addis Ababa is of 7 buses in the regional city of Jimma. Logically the enterprise (minus Jimma) should be transferred to the City administration of Addis Ababa, and this option has been explored for at least the past 5 years. However complicating factors relating to its balance sheet and certain intra-government liabilities have prevented progress in this area.

Existence of a private sector in other public utility sectors

At this stage of economic development, within the liberalised market economy introduced over the last decade, there has been no history of private sector involvement in other public utility sectors. Whilst the government has committed to transforming the private sector into an engine of economic growth, the latter has had difficulty in mobilising the necessary capital for involvement in the utility sectors. In addition, investment is largely constrained to local businesses and there has been no international involvement.

Local government policy stance and willingness to reform

The City administration in Addis Ababa has expressed a willingness to introduce reforms in the urban passenger transport sector and has been actively involved in the study on improving transport in the city.

The Transport Authority, which is the administrative organ responsible for urban transport, has accepted the recommendations of the study on improving transport and has prepared a plan for tighter self-regulation of the taxi sector. This is based on empowering area-based owners associations within the sector, and mandating membership of one of these as a pre-condition for operating the relevant services. Infrastructure service providers, such as station ‘marshals’, will be made subservient to the owners associations and it is anticipated that an orderly market will emerge.

How these reforms of the taxi sector will be integrated with those proposed for the bus sector is rather less clear. Even if competition can be attracted, there is a danger of a lack of service integration between buses and taxis. However the price differential between these two modes, currently more than 5 to 1, means that they are not really competing in the same market. The fact that taxis are providing public transport services to the extent that they do is a reflection of the lack of capacity within the city bus system.

Economic conditions

National GDP per capita

In 2000, the GDP of Ethiopia in constant 1995 US Dollars was 7.45 billion. With a population of some 63 million, this equated to a GDP per capita of US$ 118. This makes Ethiopia one of the poorest countries in the world, even when comparing with $ 617 across the whole of sub-Saharan Africa.

Over the decade from 1991 to 2000 the Ethiopian economy has grown by just over 4% per annum in real terms, though population growth has reduced its impact to some 2% per capita. However this period included a major war with neighbouring Eritrea and a significant drought that depressed agricultural output. More recent data suggests a growth rate approaching 10% in real terms (7.5% per capita) as the benefits of economic reforms begin to bear fruit.

It should also be noted that a high proportion of the population still survives on subsistence agriculture, and their non-cash output is not captured in the GDP statistics. As a result, GDP per capita will be far higher in the city than for the nation as a whole. Recent estimates put the figure as high as US$ 418, or some 3.5 times the national average.

Percentage of income spent on public transport

No reliable data are yet available on the transport proportion of household expenditure. The urban transport study has conducted an interview programme, but its findings have yet to be published. Other research has highlighted a wide disparity of incomes depending on the degree of formal employment in the household. Thus any analysis will demonstrate a range of affordability ratios rather than a typical value.

Basic salaries might be in the order of Birr 400 per month for the semi-skilled employee who might be expected to use public transport. If he lives within 10 kilometres of the city centre, and is able to complete his journey on a single boarding, then he is likely to face fares between Birr 0.35 and Birr 0.50. Taking a weighted average fare rate, and assuming 22 days worked per month, his commuting cost will amount to some Birr 20. This represents only 5% of income, and would be very low by developing country standards. Clearly this figure would increase according to any interchanges that needed to be made, especially given the current tariff structure with flat fares for each route.

If he were to be living closer to the city centre, and had to commute by taxi because of the lack of service accessibility, then his typical fare might be Birr 1.60 including an interchange (forced or desired). For the same working month, his commuting costs now amount to Birr 70 representing 17.5% of income. This figure is clearly far less affordable, but is more typical of developing country cities in general. It should also be noted that there is no intermediate level of service that might strike a more appropriate balance between quality and affordability.

Current transport fares and costs, in current dollar equivalents

Tariff systems vary between buses and taxis. For the former there is an approved fare for each route irrespective of the length of the individual boarding made on it. This fare is correlated to, but not directly dependent on, the length of the route. Variances will occur depending on how recently the route has been introduced or revised, or whether the services are nominally limited-stop.

Taking a weighted average of fares applicable to the core network, fare rates approximate to Birr 0.046 per kilometre for those riding the full length of the route and proportionately more for those making shorter boardings. At current exchange rates, this fare rate equates to 0.53 US cents per kilometre representing some of the cheapest travel in the world. Fares are so low both because of the subsidy provided to users (roughly 25% of the fare actually paid), and the extremely high average load factor (over 100%) representing tolerance of crush conditions in the peak.

By contrast, approved fares for taxis are directly related to the length of the trip being fixed within bands of kilometres run. However the setting of the band limits has resulted in certain anomalies with regard to fare rates, with these being far higher for trips up to 4 kilometres in comparison with those of 6 kilometres or more. As a result, most operators artificially break their routes (especially at peak times) so as to force passengers to pay two higher rated fares.

Taking that higher rate as being applicable to the large majority of journeys, taxi fare rates work out at about Birr 0.26 per kilometre. At current exchange rates, this fare rate equates to some 3.0 US cents per kilometre. This figure is somewhat higher than that generally typical in sub-Saharan Africa, and reflects the small authorised capacity of the vehicles being used and the good control of overloading.

Calculating transport costs is more problematic. The official accounts for Anbessa do not reflect the full true costs of its operation, but do include a depreciation charge relating largely to capital expenditures not actually made by the enterprise at a rate set by tax legislation and not economic life considerations. Notwithstanding these reservations, recorded costs are some Birr 5.3 per kilometre, or Birr 0.053 per passenger kilometre at an average load factor of 100. The latter figure represents 0.61 US cents at current exchange rates.

For taxis, their operating margins are hard to determine in such a dispersed ownership system. However attempts have been made to estimate these in Section 5.4 of this report, leading to figures of an 8% return on sales for the driver and an 8% to 15% return on sales for the owner. Taking a mid figure for the owner, total operating costs then represent some Birr 0.21 per passenger kilometre. This figure equates to 2.4 US cents per kilometre.

Population of Addis Ababa

The population of Addis Ababa was estimated at some 2.57 million in 2001/2. At an assumed growth rate of 3% per annum, this would equate to 2.81 million at the end of 2004.

An alternative source has placed the current population of Addis Ababa at 3.36 million, but this figure is not supported by a credible time series (related to incomes, for example).

Current public transport patronage

Passenger numbers by category

The formal bus operator in Addis Ababa, Anbessa City Bus Service Enterprise, maintains detailed records of its passenger carryings on each of the 89 routes that it operates derived from single-trip ticket sales. In the latest month these totalled just over 17.0 million, or some 567,000 per day. However this month included 5 Sundays and a Public Holiday, and typical carryings on a weekday are nearer to 600,000. Fare evasion is considered to be low (between 5% and 10%), giving an estimated total carryings of 650,000 per day.

In addition to the conventional adult single trip tickets, students and scholars purchase 10-trip tickets at a discounted fare valid only on school days whose use is not then allocated back to the routes. Sales of these tickets vary according to the school year, but can reach a peak of 660,000 trips in a month or 30,000 per day. Fare evasion by students is considered to be a more serious problem than for adults, and might indicate a maximum actual carrying of some 40,000 per day. Taken together, total carryings by Anbessa can thus approach 700,000 trips per day.

The informal sector makes no attempt to collect passenger numbers, only being interested in the revenue that they generate. Even then, marginal revenues in excess of the identifiable costs of operation that the driver must cover can only be a matter for conjecture. Numbers can only be estimated from constructing vehicle productivity based on driver interviews, or from measuring modal shares in comparison with Anbessa.

The study on improving urban transport carried out detailed loading surveys at a limited number of sites, from which it concluded that carryings on minibuses and modified taxis were roughly at three-quarters of the level of those on the large buses. However the average trip length on the smaller vehicles is considered to be significantly less than for the buses, and might therefore result in passenger numbers being up to 50% more than on Anbessa.

The urban transport study attempted to estimate passenger numbers carried by the minibuses and modified taxis, and came up with widely varying estimates for the number carried by each vehicle ranging from 132 to 312. The situation in Addis Ababa is complicated by the authorised fares structure that is discussed further in Section 5.2. In effect, this provides an incentive for drivers to break their trips at peak hours so as to charge an extra fare even if a passenger continues to ride in the same vehicle. If such artificial breaks are discounted, then the true number of passengers is likely to be nearer to the average of these estimates and would be consistent with international experience for this type of transport.

The number of minibuses and modified taxis actually operating in Addis Ababa has been the subject of considerable conjecture, and there is no consistency in the source documents. However the Transport Authority has now introduced a new computerised registration system from which it is possible to deduce that their combined number is just over 7,500 (with some 5,500 sedan taxis). This figure is much lower than previous industry estimates (ranging up to 12,500), but the new system of registration plates introduced in 2003 had resulted in total taxi registrations (including sedans) falling by 2,400. It is likely that vehicles needing major repairs have now been withdrawn from the register, and this would be consistent with anecdotal evidence of the recent decline in capital values that would make such repairs uneconomic.

If vehicles under long-term repair are no longer in the operating fleet, then it seems likely that daily availability would be in the order of 70% resulting in some 5,300 vehicles actually being on the road. On the basis of these combined assumptions, daily carryings would then be in the order of 1.2 million passengers per day. This figure is consistent with the industry estimates reported by the study on improving urban transport, and broadly with the modal share estimates reported above.

Taken together, these figures suggest a daily public transport patronage of some 1.9 million passengers, or roughly 0.65 trips per day per person living in the city. Whilst this figure might appear to be high, it must be recognised that there is a fair level of commuting into the city from surrounding settlements, and that the high proportion of journeys involving one or more interchanges between services results in multiple trip counting.

Ticket types

As noted from the estimation of passenger numbers above, Anbessa predominantly offers adult single-trip tickets but also 10-trip student tickets.

The single tickets are pre-printed and both batch and serial-numbered in the denominations of the applicable fares for each of the lines. Two colours of ticket are used, for the outbound and return trips, so as to assist with revenue protection and prevent immediate re-use of a ticket. The conductor completes a waybill for each trip, enabling the allocation of the relevant passenger numbers.

A static conductor, positioned immediately ahead of the rear entry door onto the bus, sells the tickets. At busy times, and especially at the terminals, ticket sales transactions are carried out through the window of the bus and the passenger is only allowed to board once in possession of a valid ticket. This procedure is considered to enhance revenue integrity, but does result in extended dwell times at stops and hence contributes to the slow operating speed. No tickets are pre-sold to passengers waiting in the queue, apparently to ensure the use of tickets from only one serial number batch on any one boarding and hence facilitate ticket inspection.

The student multi-trip ticket takes the form of a small sheet with ten tear-off strips. Each of these has days, hours and quarter hours marked, and cancelling of the ticket is supposed to involve the relevant details being marked so as to avoid re-use. When a new strip is cancelled, the previous strip is torn off and discarded. However it is reported that the driver, responsible for the cancellations, often cannot spare the time to make the three marks accurately. Also, anecdotal evidence suggests that users find it relatively easy to tamper with pen cancellations, and sometimes pre-treat the tickets to make this easier. Thus unauthorised reuse of tickets cannot be discounted, and the lack of recording against the waybill makes ticket inspection even more difficult.

Anbessa is now introducing multi-trip adult tickets, but not yet with any purchase discount so as to encourage their adoption. Apparently the only immediate benefit will be that holders of such tickets can avoid the queue for payment at the rear entry door. Clearly, a wider adoption of these ticket types requires a resolution of the fare evasion potential that they create.

Passenger load factors

The average number of tickets sold by Anbessa per trip in the last month was 104. Allowing for fare evasion, the number of adult passengers rises to 112. Adding students at the relevant times of year could raise this to 119. Not all passengers will have travelled the full length of each route, though, and the average occupancy might only amount to some 110.

However this figure is still well in excess of the registered capacity of 100 passengers per bus, and shows the degree of over-loading that is commonplace at busy times. Loads in excess of 150 people have been counted, and crush conditions have to be endured by the passengers at these times.

Part of the explanation for the very high load factor is the bi-directional nature of traffic that is observed even at the peak hours. The surveys for the study on improving transport found that the most unbalanced route had nearly half of the peak flow in the contra-peak direction, and the most balanced had less than 5% differentiation.

Anbessa is also able to maintain high load factors throughout the day, despite withdrawing very few buses after the peak, because of its strong price advantage compared to the minibus and modified taxis. This results in an extreme peaking of the minibus demand, with the result that large numbers of vehicles can be observed parked up at terminals between the peaks.

Average load factors on the minibuses and modified taxis follow very much the same pattern as for Anbessa. The study on improving urban transport identified an average occupancy of 10.8 in the peak, dropping only to 10.0 in the off-peak. These represent load factors of 90% and 83% respectively against the registered capacity of 12 persons (including the driver) for these classes of vehicle, with the passenger load factor being slightly lower when the crew are taken into account.

These figures must be close to the practical maximum without allowing overloading, which is extremely rare in the city centre though reportedly slightly worse in the peripheral areas. Again, the fact that traffic is available in both directions is a contributory factor, together with the operating practices of the sector designed to maximise these. These are examined further in Section 8.9 below.

Structure and organisation of urban public transport

Organisational structure

Anbessa City Bus Service Enterprise provides the conventional bus services in Addis Ababa. Although the Federal government owns this company, its operations are financially supported by the City administration that pays a subsidy for each passenger carried. However the level of such subsidies is gradually being reduced, and the formal linkage to passenger numbers has now been broken. The resulting removal of the performance incentive is undesirable, even if it allows for easier budgeting at the City administration.

Anbessa was originally a private enterprise holding an exclusive franchise for the provision of passenger transport services in the city, but was nationalised in 1974. It then formed one part of the Passenger Transport Corporation, with two other divisions being responsible for long-distance bus services and bus materials (spare parts and bodies) supply respectively.

In 1996 PTC was broken into its three component parts, and Anbessa established again as a freestanding commercial enterprise. Since then the workshops component of Abay Technical Services (the bus materials supply division of PTC) has been merged with Anbessa, with which it shares the Yeka depot site. This latter business provides services both to Anbessa and to the external market.

Anbessa has an explicit remit to act in a commercial manner, but it doesn’t have the freedom to set its own tariffs. These have effectively been frozen since 1992 in the interests of travel affordability by citizens, a key priority for the City administration. As a consequence, the city and national governments have provided the necessary financial and capital support to ensure both its continuing operation and its expansion to meet growing demand.

In 2000 the Ethiopian Privatisation Agency earmarked Anbessa for divestiture, but the only likely acquirer of the business under its current establishment is the City administration. In preparation for the divestiture, financial due diligence was carried by external auditors and a number of recommendations were made. These related in particular to a restructuring of the balance sheet so as to leave historic liabilities and obsolete assets with the Trustee Board for PTC. As most of these will relate to intra-government transactions, there seems little point in not liberating the enterprise from these. Unfortunately these recommendations have only been partially acted on, with the result that the enterprise remains encumbered.

Legal basis for organisation and jurisdiction

The market for passenger transport services was deregulated by the Transitional Government through the ‘Proclamation to Provide for the Regulation of Road Transport’ (No.14 of 1992). Prior to that date, minibus ‘taxi’ services were regulated on a zonal (Ketena) basis within the city. A licence had to be obtained to operate services within one of five designated areas, and the number of such licences was determined by administrative control – ‘quantity licensing’. The right to operate large buses was held on an exclusive basis by Anbessa.

After deregulation, conditions for market entry were limited to proven roadworthiness of the vehicle and qualification of the driver – ‘quality licensing’. An intending operator also has to secure a business permit from the City administration, but this is only a formality if the other conditions have been met. Responsibility for the business permits falls under the Bureau for Trade and Industry Development, but has been delegated to the Transport Authority for administrative convenience. The current cost of a permit is Birr 110 for initial documentation, plus an annual renewal fee of Birr 60. The resultant permit is valid throughout the city, and places no conditions with regards to routes or areas of operation.

The proclamation provides encouragement for the taxi sector to organise itself on an area basis, but doesn’t make this a license condition. However a limit has been set on the number of vehicles (originally 500, now 750) that can belong to any one association so as to prevent the emergence of the monopoly powers that are apparent in some other jurisdictions (Ghana, Uganda). Currently three associations are registered with the Transport Authority, but their total membership represents only a relatively small proportion of the industry. Nevertheless it is these bodies with which the Authority negotiates over all regulatory matters, particularly the determination of tariffs.

In theory, liberalisation of the urban passenger transport market also applies to conventional bus services as well. Thus Anbessa can no longer be considered to be holding an exclusive franchise within the City, but no competitor has been able to emerge to take on the challenge. Clearly the preferential access to subsidy and public investment held by Anbessa act as a barrier to any new market entry. Reportedly, also, there has been a lack of encouragement from the Transport Authority in respect of licence modalities.

However liberalisation does allow the various long-distance buses to provide some peak supplementary services when vehicles are available between their scheduled departures, and these provide useful additional capacity at these times. This is a purely opportunistic supply, though, and cannot be integrated into the formal network.

Vehicle ownership and management

Within the formal sector, Anbessa is regarded as the legal owner of its operating assets, though there is a lack of clarity regarding those procured for the former Passenger Transport Corporation before the establishment of Anbessa as a separate legal entity in 1996. None of these have been obtained on operating leases, whereby ownership rights would still be vested with the lessor.

Theoretically, Anbessa has full discretion with regards to the management of its assets and the uses to which these are put as it as been established as a commercial enterprise. However the Board of Anbessa is appointed by the Federal public enterprises supervisory authority, and the Chairman of the enterprise is the Addis Ababa City Manger. As such, Anbessa is subject to a strong level of political direction – particularly with regard to affordability of its services. This was most apparent when fuel prices were rising over the last year, as public pronouncements were made that Anbessa fares would not be adjusted accordingly despite increases being granted for the minibuses.

Vehicle ownership within the informal sector is covered in Section 6.1 below.

Vehicle purchase, finance and insurance

The fleet currently owned by Anbessa has been acquired through a number of different routes, described separately in the following paragraphs, but fortunately there has been a high degree of standardisation in their technical specification over recent years.

Excluding the 7 obsolete Fiat buses operated in the remote city of Jimma, the oldest buses in the Anbessa fleet are 138 Mercedes 1621 buses procured between 1986 and 1992 by the former Passenger Transport Corporation. Of these, at least 70 are beyond economic repair but some 65 remain serviceable for light duties such as contracts. It is planned that 60 of these will now receive a capital rehabilitation using a German government grant for the chassis spare parts.

In 1996 the first tranche of 100 DAF TD2105 buses with Belgian bodywork was acquired. This was financed by a Dutch government grant for 60% of the value, with the balance being paid by the Federal government.

In 1997 a second tranche of 163 similar buses were acquired again with a Dutch government grant for 60% of their value. The balance was funded through medium term finance from a local commercial bank, and this has now been repaid.

In 2002 a further tranche of 50 similar buses were acquired, this time using concessional finance from the Belgian government. A 25% deposit was required, paid from the resources of the enterprises, but the balance will be repaid over 20 years after a 10-year grace period. There are therefore no immediate demands on the enterprise in respect of these vehicles.

In 2003 a final tranche of 150 similar buses were acquired, using a similar approach to the first tranche but with the Dutch grant component now reduced to 50% of the value.

The local equivalent price of these buses has risen from about Birr 1.5 million to some Birr 1.7 million over these period, principally because of the appreciation of the Dutch / Belgian currencies (latterly the Euro). The US equivalent value of the last tranche is nearly $200,000, and reflects the impact of importing fully-built buses with European bodywork, as well as duty and tax totalling 28%. It could be argued that the beneficiaries of the grant component are in fact the two Belgian bodybuilders (Berkhof and Jonkheere) and not the enterprise.

Anbessa has always carried a full range of insurance covering third-party and passenger risks in respect of its operations, and property and employer’s liability risks with regard to its other assets and its employees. Unfortunately this cover has to be underwritten by the national insurance company, as it would almost certainly be more economical to self-insure for all but catastrophic damages. The cost of the cover is now equivalent to about 2% of gross revenues, and so is not insignificant in respect of the finances of the enterprise.

Issues related to vehicle finance for the informal sector are covered in Section 6.3 below.

Historically there had been no legal requirement for operators to hold an insurance policy covering third-party and passenger risks. However the Transport Authority is now planning to introduce this requirement, but the enabling legislative instrument has yet to be promulgated.

Legislative instruments

In addition to the regulatory framework described in Section 4.2 above, operation of all road vehicles in Addis Ababa is governed by the Road Traffic Safety Regulations (No.5 of 1998). These regulations define a wide range of traffic offences, the powers of the Police to enforce these, and the penalties that may be imposed for their transgression.

One encouraging feature of the regulations is that there is a system of escalating penalties for repeat offences, resulting in the suspension of licences or their eventual cancellation, that should provide a cumulative deterrent effect. Unfortunately, though, this escalation does not place sufficient emphasis on the severity of the six tiers of individual offences identified in assessing the final penalty, and the adoption of a points system would both be more equitable as well as more effective.

It should also be noted that existing instruments do not provide for the control of operators or of owners associations. This lack of collective responsibility makes enforcement measures against drivers in particular somewhat ineffective. The Transport Authority is planning to rectify this situation by strengthening the self-regulatory powers of the owners associations, and introducing formal contracts of employment for the drivers.

Staffing and skills analysis

Anbessa employs a total staff of 3,150, equivalent to 6.0 staff per serviceable bus. The ratio is not quite so impressive when measured against the number of vehicles actually deployed in commercial service daily, where it rises to nearly 8.0.

Amongst the reasons for the relatively high staff:bus ratio are the means of revenue collection using on-board conductors, and the very high proportion of all-day services provided. Only the 40 contract services and some 20 other duties are operated using split shifts.

On the other hand the enterprise is currently short of its establishment in a number of key areas, and a significant level of overtime has to be worked. This is particularly pertinent with regard to drivers, where Sunday duties largely have to be covered in this manner. Given that the standard week for transport workers is already 48 hours (as against 40 hours in non-essential services), this is not only socially divisive but also potentially dangerous.

Staff turnover is reported as being high amongst drivers, with employees being tempted away by the higher salaries and benefits available in road haulage. However the impact of their working conditions should not be ignored in this regard. Conductor turnover is also quite high as a result of disciplinary enforcement, and recently buses have occasionally been grounded by the lack of an available conductor.

Anbessa management recognise that the administrative staff overhead is excessive, but this is partly the result of inefficient systems as well as the lack of computerisation and of properly qualified staff. Development of management information systems is currently a high priority.

Skills levels are not considered a major problem at the operative level, but it is recognised that the enterprise has difficulty in attracting and retaining good quality supervisors and other managerial staff. Although Anbessa is not restricted by government pay scales, its increasing lack of profitability does restrict management’s freedom of action to pay full market related salaries. This has particular implications in the finance department, where there is only one qualified accountant, and also in the engineering department.

Within the informal sector, driving standards and vehicle condition were both indicative of a low skills base in the sector. The Transport Authority is aware of these problems, and is now putting programmes in place to provide the necessary incentives for their enhancement.

With regard to driving standards, it is now recognised that enforcement is being weakened by the relative ease with which licences can be forged. The laminated pouches within which these are sealed can be split open, and photographs and other details then altered. New security systems have been devised, and tenders issued for the necessary equipment.

Vehicle standards are expected to be raised through the very recent addition of private-sector testing capacity, but it is still too early to identify the impact that this measure will have. However the contractual relationship between the Transport Authority and the testing stations does not appear to include adequate performance incentives, or penalties for transgressions of their responsibilities.

Facilities and equipment

Anbessa operates out of three sites, its headquarters complex at Yeka to the east of the city, a second depot at Shegole to the northwest, and a newly opened depot at Mekanisa to the south.

The Yeka depot is 7.0 hectares in area, excluding the bus-building workshops complex of Abay Technical Services that is immediately adjoining. The operating depot has the facilities for 300 buses to be based there, but currently only 221 are deployed. The site also includes the central workshops for the enterprise (developed with Dutch technical assistance), and the main spare parts warehouse. The headquarters office provides both for central administrative and depot control functions.

This depot is sited immediately adjacent to the recently completed city ring road, providing for effective vehicle positioning on routes. The depot yard is surfaced, and well laid out for efficient operation. The workshops are well equipped, and housekeeping standards are high. The fuel station has adequate capacity for the fleet based on the site. No significant near term investment is required.

The second depot at Shegole is 5.4 hectares in area, but the yard is not surfaced. This facility has its own fuel station, and light maintenance and running repair facilities as well as a spare parts store operated on a satellite basis from Yeka. 300 buses could be based here, but at present only 171 are deployed.

This depot is self-sufficient apart from heavy maintenance that is still carried out at Yeka, leading to some loss of control. Operating standards would be improved by surfacing the yard, but it is understood that the necessary finance has not been made available. Otherwise no major near term investment is required.

The third depot at Mekanisa has only recently been made operational. It has a site area of 7.3 hectares, which would be more than sufficient for a further 300 buses; however only 73 are deployed there at present. Facilities are still being developed, but it is intended that this site is made increasingly self-sufficient. The major outstanding investment need is for the yard to be surfaced.

Financial environment

Anbessa operates in an uncertain financial environment. As a commercial undertaking, it is expected to act on sound business principles. However it lacks the freedom to set its prices to maximise its income, or even to recover the increased unit costs of its essential inputs. It is not able to make adequate provision for the replacement of its assets from its operational cashflows, and its business expansion has almost entirely been externally funded.

In recent years it has been unable to recover its cash expenditure from its commercial income, and has relied on subsidy paid by the city of Addis Ababa to cover the shortfall. However this subsidy is being decreased year on year, and the city has committed to its eventual elimination. Neither the owner of the enterprise (the Federal government) nor its manager (the City administration) has made any consistent commitment to its long-term capital expenditure programme.

Despite the near freeze on tariffs, service income has grown as a result of the expanding fleet. Operating expenditure has increased for the same reasons, but there have also been major impacts from a wage restructuring in the late 1990s, increasing fuel prices in the last couple of years, and higher maintenance costs as the original DAF buses start to age. However fixed cost recovery has improved as business activity levels have increased. As a result, although recoveries of cash from direct enterprise earnings before subsidy have been negative, their level has been broadly constant over the period under review.

When the fleet depreciation charge is included, Anbessa has become increasingly unprofitable over the last three years. However depreciation is a charge legally computed against the full market value of the assets, irrespective of how these were acquired or financed, over a period set externally. It could be argued that this approach is correct if Anbessa were being required to provide for its asset replacement from its internal resources (amortisation), but the recent history of the enterprise does not create such an expectation. Also the period of depreciation, recently set at 8 years, is probably too short in relation to the economic life of the assets; 12 years would be more appropriate.

Once the depreciation charge is added back, Anbessa is seen to have been generating a strong cashflow in the eight years since it was revitalised by the first of the new DAF buses. As a result, Anbessa has discharged its short-term liabilities and so currently incurs no interest charges (except for the Belgian soft finance, which incorporates a 10-year grace period).

The impact of these various factors is shown in the following table:

| |  |1996/97 |1997/98 |1998/99 |1999/00 |2000/01 |2001/02 |2002/03 |2003/04 |
| | Subsidy |25,894 |45,153 |50,089 |49,134 |42,972 |36,645 |23,104 |25,876 |
| |Expenditure, of which |60,067 |89,755 |108,106 |117,144 |121,162 |124,437 |141,461 |169,402 |
| | Depreciation |26,525 |39,070 |42,208 |41,377 |39,108 |39,621 |43,100 |59,059 |
| |Cash recovery |-2,818 |-2,250 |-2,033 |-4,659 |-1,458 |-4,466 |-866 |-5,335 |
| |Profit |-3,449 |3,833 |5,848 |3,098 |2,406 |-7,442 |-20,862 |-38,518 |
| |Cashflow |23,076 |42,903 |48,056 |44,475 |41,514 |32,179 |22,238 |20,541 |

Clearly the issue of fleet replacement needs to be resolved if Anbessa is to be considered as a going concern. At the same time, a policy of freezing fares while reducing subsidy cannot be sustained forever before the enterprise starts to consume cash. It will be noted that cashflow over this period has almost entirely been generated by the subsidy income and so has declined since the latter peaked in 1998/99.

At the same time its balance sheet needs to be restructured so as reflect its current status as a freestanding commercial concern. Assets and liabilities that were transferred from the defunct Passenger Transport Corporation have not been properly resolved in accordance with the recommendations made to the Ethiopian Privatisation Agency some five years ago. Unless this is done, the enterprise will not have the necessary commercial freedom to fulfil its remit.

Role of the informal sector

The informal sector in Addis Ababa provides minibus and ‘modified taxi’ services throughout the city. The latter vehicles are pick-ups converted to passenger carriage by the addition of a canopy over their load bed, fitted with bench seats and windows on each side. Their passenger capacity is the same as for the minibuses, being restricted to 11 persons other than the driver by regulation, but space and access are considerably more restricted.

As noted earlier, this sector carries over half the passengers but meets slightly less than half of travel demand when the shorter average boardings compared to buses are taken into account. These shorter average boardings arise from manipulation of the fare structure, particularly at peak times, so that longer trips are broken into shorter stages at intermediate stations. Any passenger wishing to continue with the vehicle then has to pay an additional fare.

Minibus services can be regarded as high quality in comparison with conventional buses, as their interior layout is spacious (9 seats in the main saloon, compared to a minimum of 12 in most other jurisdictions) and over-loading is rare over most of the network. However the condition of the seats and cleanliness of the interior often leave much to be desired.

Modified taxis tend to operate on poorer and unmade roads, and serve market areas where the carriage of produce as well as people is important. Their seating is much more cramped than for the minibuses, and the carriage of produce often affects odour and cleanliness. They are also generally older than the minibuses, having been developed locally to meet emerging demand at a time when minibus imports were not possible.

Fares levels are in the order of four times those charged by Anbessa, at least for the lengths of trips typically undertaken. Clearly the market for such services is greatly reduced for any discretionary travel made outside the peak hours, and this explains both the lower fare rates available at these times and the low level of vehicle utilisation observed. Large numbers of minibuses can be seen parked up at the main stations during the day.

As a result, the informal sector can be seen as a premium peak service supplement to the core all-day passenger services provided by the city bus operator. However their aggregate volume of supply is in the same order as the core service. Overall network efficiency would suggest that these two functions should be differentiated, and each then optimised to its task.

Ideally the private sector should provide both a commercial bus service complementing the subsidised network of Anbessa, and a range of premium products designed to segment the public transport market and proving capable of attracting car users. Existing minibuses are too small for the former role, and may not be sufficiently comfortable for the second.

Private investment could be attracted to the sector if the appropriate regulatory and enabling framework were put in place. Unfortunately, though, this has not been encouraged so far.

Service supply characteristics

Operational characteristics – bus services

A description of the operational characteristics of Anbessa City Bus Service Enterprise for the month of Tir 1997 EC (9 Jan to 7 Feb 2005) is attached as Appendix 1 to this report. The following commentary is based on an analysis of these data, together with interviews held with the senior management of the enterprise.

Routes operated

At the current time Anbessa operates 89 of its 93 designated routes, with Routes 1, 22, 86 and 91 being suspended. On all of these routes apart from six (Routes 7, 58, 69, 78, 88 and 89), services are run throughout the operating day, nominally from 06.00 to 22.00 hrs. 9 of these routes are peri-urban, connecting settlements outside the city boundary that fall within the travel-to-work area. The balance provide services within the conurbation, mostly on radial routes to the central business and commercial areas of the city.

On 25 of these routes there are supplementary peak services operated according to vehicle availability, with only 6 of these receiving a regular vehicle allocation. The peak services are usually run over a shorter portion of the designated route, but still charge the same fare as for the full service. In this manner, they help to provide a more accurate match of transport supply to passenger demand.

A high proportion of the routes are operated to the three main terminals of Anbessa: Addis Ketema (Merkato) (31 routes); Leghar (22 routes); and Menelik Square (9 routes). Another 7 interchanges are also used, covering 26 routes, but most routes originate from street-side facilities in the suburbs.

Anbessa was not able to provide a route map, but it is immediately apparent that there is a high degree of overlap between routes within the network. This primarily arises from the fare structure, whereby the designated fare depends on the length of the particular route and not on the length of an individual boarding. It is therefore in the interests of the enterprise to provide services in the corridor that can command higher fares, and not to integrate these with lower fare and shorter services.

Scheduled timetables

All Anbessa services are operated to scheduled timetables, but no information about these is provided at the large majority of stops. Headways on many city routes are extended, and service intervals often fail to provide clock-face regularity.

Only 5 routes provide a service frequency of 6 per hour or better, the normal standard for an urban bus service. A further 26 routes provide 3 or 4 services per hour, but 31 only offer two services and 18 only one. The nine peri-urban services are operated at headways ranging from one half to three hours.

Scheduled running times on all routes are the same for each trip, irrespective of the time of day or the direction of travel. The latter factor is of particular significance in Addis Ababa because the hilly terrain in the northern half of the city results in lower traffic speeds in the uphill direction. A computerised bus and crew scheduling system would be able to adapt to these variances, but this is beyond the scope of the existing manual procedures.

As a result, trips are often lost in the peak when buses cannot keep to their scheduled time. Lost trips in the month averaged 960 per day, and will have exceeded 1,000 on most weekdays. Lost trips totalled 15% of all those scheduled in the month, a deterioration from the 11% reported for three months ago in the urban transport study.

Details of vehicle kilometres run

Anbessa has a peak vehicle requirement for its core services of 344 buses. Some 1.82 million kilometres were run on these in the month, representing 176 kilometres per bus per day. A further 31,500 kilometres were run on peak supplements, a negligible 1.7% of the total.

On six routes bus utilisation was less than 100 kilometres per day, with the worst achieving only 45 kilometres (though this was particularly badly affected by lost trips). Given the high level of passenger demand, this would appear to represent a wasteful deployment of buses.

Total kilometres reported for 2003/04 by the urban transport study were 21.6 million. It should be noted that this is only an increase of 15% over the 18.8 million kilometres that were achieved in 1998/99 despite the injection of 200 new buses in the intervening period. Part of the problem may have been a decline in operating speed, though new road infrastructure should have helped to offset the congestion effects of increasing motorisation. However the main concern must be the declining operational availability of the 1996/97 DAFs, as well as the use of some of these on contract services.

Driver training and control

Anbessa suffers a high level of turnover of its driving staff, resulting in the need for intensive training of new recruits. However observed driving standards are high, and driver discipline was reported as good in the two referenced studies. Supervision and training standards were both also assessed as good.

Accidents have averaged some 164 per month in the second half of 2004, with 5 fatalities occurring in that period. An accident rate of 8.9 per 100,000 kilometres must be considered high, but is more a reflection of general chaotic driving standards in the city than those of the enterprise.

A more detailed analysis for two months suggested that drivers had been responsible for 41% of accidents. However the large majority had resulted in little or no damage, with the more serious accidents (damage above Birr 1,000) accounting for only 6% of the total.

Vehicle maintenance practices and service standards

Anbessa operates a complex preventive maintenance programme, combining both time and distance related interventions. Buses are inspected both weekly and monthly, and oils are changed at 7,500km (engine) and 25,000km (transmission). No more intensive interventions were identified other than the preparation for annual roadworthiness inspection.

It was also noted by the urban transport study that a lack of co-ordination between the operational and technical planning functions meant that kilometre data for service scheduling were often received too late for integration with other preventive maintenance activities. This resulted in a loss of accuracy, and additional disruption to operations.

To avoid such problems, it is normal practice in urban transport undertakings to plan all preventive maintenance interventions on a time basis. This would be readily applicable in Addis Ababa as the large majority of buses work on all-day duties, and each bus is rotated through the full range of depot duties such that kilometre allocations are equalised over time.

An analysis conducted for the study identified that 2,004 interventions had been scheduled in November 2004, but that only 1,420 inspections had been achieved and work then carried out on only 1,226 occasions. This represents a performance ratio of 71% for inspections, and 61% for interventions.

Clearly this level of performance is deeply unsatisfactory. The General Manager reports that operational availability is given a higher priority than adherence to maintenance schedules, and this inevitably results in a deteriorating technical performance. It would be strongly recommended that such a policy be reversed immediately. If integrated with a revised time-based plan for preventive maintenance, this need not result in any loss in bus utilisation.

A limited survey of daily non-availability for the front-line DAF fleet showed an average of 78 buses suffering from technical defects, of which 63 were under long term repair. This represents a technical availability of only 83% for a fleet averaging little over 5 years in age. Given the proven durability of the vehicle specification in respect of over-loading and harsh operating conditions, this must be regarded as an unacceptable performance.

Reliability is more difficult to determine from readily available data. The study on improving transport reported some 135 ‘breakdowns’ per day. However this term is used by Anbessa for any technical intervention, even a quick brake adjustment carried out at a terminal, and so does not automatically imply an unscheduled service interruption.

The enterprise also records any occasion when the bus assigned to a route has to be changed. However this is just as likely to be for operational reasons, arising from the delayed arrival of the designated bus and the need to clear passenger queues, as it is for a return to garage. Thus the number of bus changes cannot be taken as a proxy for breakdowns either.

Any quantification of reliability statistics would require a new format of reporting for the data that is being collected at various points in the control systems. Such data would normally be presented in relation to the severity of its impact on the operation, as follow:

← stoppage on the route, requiring on-site repair or tow to garage;
← return of bus to garage for attention to significant defect;
← defect report rectified by field maintenance at terminals; and
← defect report made on the return of the bus to garage at the end of its duty.

Service monitoring

Service performance, particularly in relation to schedule and trip achievement, is carried out at the three main terminals and at a further 17 control points in the city designed to capture the large majority of the routes being operated. Controllers at these points also monitor the correct issuance of tickets to passengers.

The controllers send a daily control chart to the Quality Control Department giving both the scheduled and actual time of operation of each trip. These data are collated and analysed so as to inform operational management.

Bus terminal management system

Primary control of the route network is carried out at the three main terminals, Addis Ketema, Leghar and Menelik Square. Terminal controllers adjust services as required to balance the implications of late arrivals and lost trips, and re-allocate vehicles accordingly in order to meet the most pressing levels of demand.

Technicians are also based at the terminals, and carried out some 110 interventions per day in the month of November 2004 in order to prevent the need for buses to be returned to depot.

The terminals are congested, and poorly laid out to deal with the volume of bus movements and the high number of routes being operated through them. The study on improving transport recommends that any development of their facilities should be delayed pending restructuring of the route network in response to a scientific travel demand study. Revisions arising from this would probably result in more routes being run through the city centre, and less actually being terminated in it.

Operational characteristics – minibus services

Routes operated

The urban transport study identified 106 minibus routes in Addis Ababa. However there is clear evidence that the operators break these routes down into smaller sections, particularly at peak times, in order to exploit the authorised fares structure and maximise their income. The Transport Authority estimates that there may be as many as 300 route segments, but has not carried out a detailed analysis.

There are two consequences of this multiplicity of routes within the urban network. First the level of service that can be justified on each is either infrequent, raising passenger waiting time, or requires the use of small buses that are relatively costly to operate and are inefficient users of road space. Second, each route is likely to have a high overlay on other routes that increases the potential for on-road competition even within a regulated system.

The Transport Authority is planning regulatory reform for this sector, but this will not take the form of route franchising or service contracting. Rather the development of strong area-based owner associations will be fostered, and these will take responsibility for the design and operation of routes in their zone.

Scheduled timetables

Minibus taxis do not operate to scheduled timetables. Rather, the practice in the industry is for each vehicle to wait in line at its terminal until has been filled (‘fill and run’) and only then to commence its operation. This can provide a good level of service throughout the day on more busy routes, but often results in excessive waiting time in the off-peak period for quieter routes.

Service accessibility along the routes also suffers, as vehicles passing intending passengers near to the terminals will mostly already be full. As a result, passengers are forced to walk to the terminal even if another boarding point would be more convenient for them.

Vehicle kilometres run

Obtaining estimates for minibus kilometres is not at all easy. Owners have no clear idea of what their drivers actually achieve each day, and most vehicles don’t have working odometers that would enable them to monitor this. The drivers don’t think in terms of kilometres, but rather in terms of the fuel they use (litres) or the cost they incur (Birr). Both of these figures need to be interpreted through the different prices for petrol and diesel, and assumed rates of consumption under Addis Ababa operating conditions.

The study on improving transport interviewed both the owner associations and a number of drivers at three of the main terminals (Piazza, Merkato, and Saris). Owners believed that it was possible to cover on average 200 kilometres per day on ‘hot’ routes, and that this would involve 6 to 8 round trips in the day implying an average route length of 14 kilometres. Given that the weighted average trip length for minibus users surveyed was 5.5 kilometres, this would appear to be on the high side.

The drivers, moreover, indicated daily utilisation over the wide range of 50 to 160 kilometres, with route lengths of 6 to 10 kilometres. Recognising that the vehicles at Merkato are strongly influenced by market activity, possibly involving the carriage of produce as well as people, their data should probably be ignored. Figures from the other two sites suggest the possibility of 140 kilometres in an area with all-day demand (Piazza), and 70 kilometres in one where the demand is highly peaked (Saris).

Owners and drivers associations were more consistent in their estimations of fuel spend, with this ranging only from Birr 120 to 140 per day. This is broadly consistent with estimates given by the Transport Authority of 30 litres of fuel per day (Birr 129 and 165 for diesel and petrol respectively).

However these figures make no allowance for the very different consumption rates for diesel and petrol vehicles, which might be expected to be in the order of 8 and 5 kilometres per litre respectively under Addis Ababa operating conditions. Implied distances run therefore range from 240 to 110 kilometres. Observation suggests, though, that the large majority of minibus taxis are still petrol engined (vehicle age, reported country of sourcing, engine noise at idle, relative absence of black smoke, need for power at altitude in hilly terrain), and so the implied distance operated would be at the lower end of this range.

The more detailed driver interviews at the three terminals again gave rather lower figures than the associations’ estimates. Those at Piazza indicated a daily spend of Birr 130, which would equate to 120 kilometres at the assumed consumption for petrol-engined vehicles and this would be consistent with their kilometre estimates. Those at Saris indicated a daily spend of Birr 90, that would equate to 80 kilometres on the same basis. Perhaps the average rate of fuel consumption is slightly higher without much off-peak operation out of congested traffic.

The urban transport study attempted to reconstruct the financial viability of minibus services, and used a figure of 180 kilometres per day based on 6 round trips on a typical route of some 15 kilometres in length. However, as justified above, this assumption on route length seems to be excessive. Its derivation came from an arithmetical mean of observed route lengths, and not the geometric mean of 11 kilometres that is probably more appropriate. On the latter basis, daily kilometres would be around 135.

Fuel consumption was estimated at 25 litres of petrol daily, suggesting a distance operated of about 125 kilometres at the assumed consumption given earlier. Combining the two estimates would result in a typical daily operation of 130 kilometres. This would be broadly consistent with findings in the similarly sized city of Accra, and with the earlier analysis.

On that basis, and assuming that 80% of the 7,500 registered minibuses are put into service each day, then the kilometres covered are in the order of 800,000. For a 30-day month, this would then approximate to 22 million kilometres run (allowing for some service reduction on Sundays). This is more than ten times the kilometres run by Anbessa, and would be consistent with a broad equality in their transport volumes allowing for the difference in their respective passenger capacities.

Driver training and control

The drivers of all taxis, both sedans and minibuses, require a Class 3 licence with a higher standard of qualification and testing than for a private car. Drivers of larger buses require a Class 4 licence, with even more stringent standards. However testing in this case is likely to be on a freight vehicle that may not be as large as a full-sized bus. These requirements do provide the potential for a better standard of driving of taxis and minibuses than for their equivalent sized cars, and of large buses in comparison with smaller vehicles.

However the Transport Authority advises that the current design of driving licences can easily be tampered with, and relevant details altered or the photograph replaced. As such, a new more secure design of licence is to be introduced with tenders for the necessary equipment already having been issued.

Pending the introduction of this technology, the Transport Authority is severely hampered in its enforcement activities. In particular, the more severe range of penalties applicable to repeat offences become impossible to apply if licence details are altered.

Vehicle maintenance practices

No operators interviewed claimed to undertake any preventive maintenance for their vehicles. Rather, their general practice is to undertake essential repairs only when absolutely necessary to ensure continued operation. Requirements for roadworthiness assumed a low priority under this regime, and no real pressure was felt from the enforcement agencies.

The Transport Authority has recognised this problem, and is hoping to raise maintenance standards through more effective technical inspection to be undertaken by 18 private vehicle test stations. However examination of the contract for these services suggests that there may be insufficient performance incentives, and ineffective sanctions against transgressions of their duties.

Service monitoring

There is no body, either within the local civil administration or the operator associations, that monitors and controls service delivery on a formal basis. However the Transport Authority has a small research group that monitors services on an ad hoc basis in order to improve its understanding of the workings of the sector.

Equally, there is no authority to which a passenger can formally make complaints other than the owners’ association if the offending vehicle belonged to a member, and the association can be identified. Both situations are unlikely, with low membership rates in the industry and no specific vehicle markings. The Transport Authority will receive complaints, but again on an ad hoc basis.

The Transport Authority has recognised the problem, and is planning to introduce regulatory reforms to strengthen the powers of the owners’ associations and facilitate their effective implementation through restructuring on an area basis. Responsibility for operational control will rest with the associations, but the Authority would still monitor service delivery.

Bus terminal management

Minibus terminal management is provided by self-appointed tera askebari, literally taxi-order attendants. These are referred to in the referenced studies as ‘marshals’, and that term will be used throughout this report.

The urban transport study reports that the marshals have formed into informal ‘community co-operatives’ or ‘unions’ based on terminals. Membership fees are reputed to be as high as Birr 2,000 per year, which is both indicative of the potential earnings of the marshals and of the sums of money flowing through the system.

It seems most unlikely that the co-operatives would need to collect such sums for their own immediate purposes, and this therefore suggests that there is probably some hidden authority beyond these organisations. Anecdotal evidence has identified individuals who appear to have benefited personally from these arrangements, but detailed investigation would be needed to quantify the problem.

Minibus drivers make payments to the marshals either for the right to use the terminal for the day if operating on a fixed route, or per departure if floating in the network. Charges are between Birr 0.5 and 1.0 per departure, and could result in payments up to Birr 10 per day as a consequence according to the urban transport study.

Some commentators think that the true figure could be even higher, when charges at the interchange points are also taken into account, and might even reach Birr 20 per day. If so, the extraction from the operating industry would be around US$ 2.3 per vehicle per day. This would still place it towards the bottom of the range falling between Accra (US$ 1.7 per day) and Kampala (US$ 10 per day).

Again the Transport Authority has recognised the problem, and plans to address this through its regulatory reforms. The proposed structure would require the marshals to be employed by, and reporting to, the owners associations. The Authority considers that it has the necessary powers to instigate such a relationship, and to enforce its implementation. It also recognises that the transition may prove problematic, but that the necessary legal powers are in place to break the hold of any unauthorised controlling bodies should that prove necessary.

Current bus routing and scheduling plan

The current bus routing and scheduling plan for Anbessa is fully covered in Appendix 1 to this report. It should be noted that this is prepared internally by Anbessa, and not externally by any regulatory or planning authority. The introduction of any ‘limited competition’ regime of economic regulation would require that this capability be transferred to the Transport Authority.

As noted earlier, there is no centralised bus routing and scheduling plan for minibuses. Their route network has developed over time to link the fixed terminals on an ad hoc basis as demand has been identified for the potential service offer. However these routes are then broken down into smaller sections, reputedly so as to provide route interchange opportunities. Drivers then often exploit these interchanges, particular at peak hours, so as to maximise their effective fare rate within the approved tariffs.

The urban transport study has identified 106 discrete routes operated by the minibuses, with the number of vehicles assigned to each being in the range between 20 and 400. In total, the number of minibuses identified is greater than the number registered for commercial service by the Transport Authority, so there must be a degree of double counting. In any case not all minibuses run on fixed routes, with many choosing to float in the network responding to perceived demand from time to time.

No routes are operated to schedule, but rather are responsive to passenger demand from their originating terminal. As passengers arrive at the terminal they are allocated to the first vehicle waiting to serve their intended route, which then commences service once it has been filled. Even on the quieter routes, there is no set maximum headway between services. This may cause severe delays to waiting passengers, both at the terminals and along the routes.

Income and expenditure statements

A 10-year history of the income and expenditure of Anbessa City Bus Service Enterprise is attached as Appendix 2 to this report. However it has to be noted that these figures have been derived from management accounts, and the last three years have not yet been audited. Further, where audits have been undertaken, large discrepancies have been observed between the management and audit figures. Comparison between the two sets of figures available suggests that those provided by management appear to be the more credible on the basis of inter-year progression.

It should also be noted that these income and expenditure statements do not provide a full disclosure of the true costs incurred by the enterprise. In particular, there is no provision for the cost of capital employed in the business. This capital, in the form of the rolling stock that has been provided to the enterprise, has largely been funded by external grant rather than internal resources. Further the enterprise has free use of its extensive depot sites and developments, though it does meet utility costs and repairs expenses in respect of these.

Despite the nature of its capital formation, Anbessa makes a charge for the depreciation of its asset base at the full open-market valuation of the cost of acquiring these. In the case of its fleet, vehicles are fully depreciated on a straight-line basis over a period of 8 years in accordance with tax legislation. However this period is considerably shorter than the likely economic life of these assets, which could be up to 15 years. Amortisation over, say, 12 years would give a truer representation of the values that should be charged to the accounts.

No balance sheet analysis has been conducted, as the official figures are encumbered by both assets (receivables) and liabilities allocated from the former Passenger Transport Corporation for which there is no supporting documentation. The financial due-diligence report made to the Ethiopian Privatisation Agency by external auditors recommended that the PTC Board of Trustees resolve these, and that Anbessa be properly restructured financially as a standalone commercial enterprise. Unfortunately, the majority of the relevant recommendations have yet to be fully implemented.

Taking these reservations into account, the income and expenditure statements do indicate the financial health of the enterprise since its revitalisation commenced in 1996 under the Dutch technical assistance programme. Prior to this time Anbessa had been trading at a loss that reached a level approaching 90% of sales in 1995/96, and was clearly insolvent. However the injection of new buses and the provision of subsidy by the City administration enabled a rapid turnaround, and the enterprise traded profitably for four years peaking at 5.1% return on sales in 1998/99.

However the City administration then committed to a gradual reduction in subsidy payments, with the intention that these eventually be phased out. With fares effectively frozen, and the unit costs of labour, fuel, and spare parts all escalating, the enterprise has since been driven into increasing losses reaching nearly 30% of sales in 2003/04. Nevertheless, when the impact of the depreciation charge is stripped out, the enterprise is still generating a strong, albeit declining, cashflow. This has enabled it to discharge all its short-term liabilities, and hence there has been no interest charge for the last four years.

Whether the enterprise is a going concern depends entirely on the security of its access to subsidy, both to cover its operational expenses and to provide for the eventual replacement of its assets. It is understood that the City administration might be prepared to commit to these responsibilities, but it is not the owner of the enterprise. However if ownership were to be transferred from the Federal administration, certain balance sheet liabilities including intra-government debts would be crystallised. Negotiations to resolve this issue are reported to be at an impasse.

No income and expenditure statements are available for the informal sector, and this is implicitly recognised by the Federal Inland Revenue in its imposition of a flat-rate income-tax levy on each operator. The owner of a minibus is assessed to be earning a profit of Birr 6,620 annually with a vehicle up to 15 years old, and Birr 5,380 for an older vehicle. The tax payable is Birr 482 and Birr 358 respectively.

Investment behaviour in the sector is suggestive of a low level of actual profitability with fares controlled at their current levels. Anecdotal evidence suggests a significant contraction in the resale value of minibuses, with these having reduced from around Birr 120,000 to some Birr 70,000 to 80,000.

Both the urban transport study and the study on improving transport have attempted to build up a picture of typical income and expenditure in the sector, but have been hampered by the lack of credible source data. This relates in particular to the typical kilometres of operation, as have been explored in Section 5.2.

From that analysis, data from the driver interviews at Piazza might be considered as being representative. Passenger income was declared at Birr 250 per day, which equates to about Birr 1.8 per kilometre run. Given that the observed load factor on minibuses is 90% in the peaks and 83% in the inter-peak, the implied fare rate per passenger kilometre is 18 cents. This figure would be on the low side if the opportunities for splitting trips were exploited, and so the revenue figure may be an (understandable) understatement.

Main expenditure heads identified were fuel (Birr 130), crew expenses (Birr 30), and payment to marshals (Birr 8). This would leave a gross margin of Birr 82, which would be insufficient to meet the reputed daily hire rate of Birr 100 far less the profit for which the driver works. This would seem to confirm that the revenue figure quoted must be understated.

The urban transport study considers that it is necessary for a minibus to collect daily revenue of Birr 300 to 325 to make its operation viable. The necessary revenue rate for 130 kilometres of operation would be Birr 2.4 per kilometre, and the implied fare rate 24 cents per kilometre. This is quite possible with careful exploitation of the fares structure, equating to Birr 1.6 being charged for 6.7 kilometres of travel.

Identified expenses are the vehicle hire (Birr 100), fuel (Birr 140), crew expenses (Birr 40), and payment to marshals (Birr 5), totalling Birr 285. This would leave a margin for the driver of Birr 27 in addition to his salary of Birr 20. These figures seem broadly credible.

From the owner’s perspective, he needs to pay the driver Birr 20 from his daily hire charge, leaving a net income of Birr 80 per day. This would equate to nearly Birr 2,000 per month if he were able to achieve 80% availability. From this he needs to deduct his fixed costs (very low, other than the assessed income tax) as well as vehicle repairs and tyre replacement. The two referenced studies have provided no credible estimates for these, but it is not implausible that they could consume up to half his gross income on average.

Were this to be the case, his annual net income would be little more than Birr 12,000. This would equate to a return on investment of 16% before depreciation if the value of the minibus were Birr 75,000, but the apparent return is highly sensitive to the assumptions made. This low earning potential would help to explain the recent drop in capital values, though it should be noted that these are still relatively high by regional standards.

It should also be noted that the Federal Inland Revenue assesses income tax on a minibus owner on the assumption of annual sales of Birr 25,000 and profit of Birr 6,620. The sales figure would be a close approximation to that derived above, but the profit estimate is rather less. If the Revenue are correct, the return on investment falls to 9%.

The minibus driver is also assessed to income tax, though this time on assumed monthly earnings of Birr 500. This would be reasonable reflection of the salary that he draws from the owner, but not of his profit margin or out-of-pocket expenses. Unless the latter were deemed an allowable expense, his true income could be nearer to Birr 1,600.

Vehicle related data

Vehicle ownership

It has not proved possible to obtain reliable information on vehicle ownership. Although this information is required on the Vehicle Registration Certificate, no database is available to search for linked ownerships or other patterns.

From discussions with the owner associations, it would appear that vehicle ownership falls into three main categories:

← owner-drivers who have procured a vehicle for their own use;

← non-transport investors, mostly in other formal employment, who own one or more vehicles as a means of supplementing their income;

← expatriate Ethiopians who may have purchased the vehicle abroad and shipped this into Addis Ababa in order to provide a means of livelihood for their relatives.

Of these categories, the non-transport investors may prove to be the most significant in terms of the number of vehicles operated. However, in many cases their commitment to the industry is not strong. Typically they will benefit from the strong cashflow that the daily hires generate, but make no reserve for heavy repair bills or eventual vehicle replacement. When faced with such expenditure, they often choose to sell up and leave the industry.

Fleet ownership is far less common in the urban passenger transport sector than in inter-city services that are seen as being more lucrative. The study on improving transport identified that the large majority of operators own only one vehicle, with very few having more than four (though one had seven).

In general, minibus owners are not able to access commercial finance to support their vehicle purchase; in any case, the cost and conditions imposed would probably be seen as deterrent. Rather, vehicles are procured either from the earnings of another business activity or through micro-finance schemes (EKUB). In the past it was reported that sedan taxi owners could trade up to minibuses, but the collapse in their net earnings as a result of fuel price increases no longer makes this likely.

Constraints on supply of sector inputs

The urban passenger transport sector operates under few direct constraints, but the nature of its operating fleet does present certain problems with regard to maintenance in particular.

Fuel is widely and readily available at a price that appears to be below the general international level. Diesel retails at Birr 4.31 (US$ 0.50) per litre and petrol at Birr 5.50 (US$ 0.64). It may be that prices are being suppressed in advance of the elections due in May 2005, and that prices might be realigned to world values after this.

For the Anbessa fleet, all recent acquisitions have been DAF TD2105 chassis bodied in Belgium by Berkhof or Jonkheere. Specification has as much been dictated by the source of the donor support, and does not necessarily reflect local market preferences. The use of imported bodywork adds both to the first cost (as bus body building is labour intensive), and to the shipping and transit cost to Addis Ababa.

Other large buses, most of which are used for long-distance services, tend to be locally bodied on imported chassis. The Iveco agent, AMCE, is most active in this regard using CKD chassis kits. There is a range of private body builders, as well as the bodybuilding workshop of Abay Technical Services recently merged with Anbessa. Despite this local capability imported bodywork can also be observed, particularly Cacciamali of Italy.

Midi-buses tend to be locally bodied light commercial chassis, particularly Isuzu. Again these are mostly used for long-distance operation, but would be equally suited to urban services as demonstrated in Nairobi. Seating capacities from 25 to 39 are possible.

However the large majority of small buses in Addis Ababa are imported second-hand, mostly from the Middle East but also from Europe. Pickups were converted locally into ‘modified taxis’ by the fitment of a glazed canopy over their load-bed and bench seats on each side. Entry is by a rear door, with the bumper acting as a step. Minibuses were either converted from light Japanese forward-control vans, or were purchased in staff-bus configuration. Because of the restriction on passenger capacity, no fold-down seats are used. Entry is by a sliding side door.

There are no restrictions placed on such imports, either in respect of their specification or age. Customs duty and tax rates are not determined in terms of the vehicle’s age, but only its seating capacity. The duty for minibuses (10-15 seats) is 35%, and that for larger buses (16 seats and above) is 10%. In each case, VAT is levied at 15% and a withholding tax of 3% is also imposed.

For operators, the real problem presented by old vehicles is being able to obtain appropriate spare parts at an acceptable quality and for an affordable price. The modified taxis may be 20 or more years old; minibuses are more typically between 10 and 15 years of age. These vehicle types have been out of production for many years, and their manufacturers are under no obligation to continue to supply parts. A parallel industry for the supply of ‘pattern’ parts has emerged, but these do not then always match up to original equipment manufacturers’ standards.

Financial arrangements for fleet procurement

Discussions with finance and leasing companies indicated only a small level of involvement in the urban passenger transport sector. Because of bad historic experience with defaults and absconding with vehicles, facilities are now only provided for operators who have service contracts to provide the necessary security. Effectively this excludes the vast majority of the industry who ply for hire and reward.

Despite the above, the leasing sector would be prepared to re-examine options for financing buses against the business case presented by service contracts. Were these to be backed up by partial-risk guarantees against regulatory failure, it would no longer be necessary to ensure positive equity in the vehicle, and hence the conditions on deposit and term could be relaxed. Interest rates would still be problematic, though, because of their own cost of funds.

The alternative for minibuses is micro-credit, as represented locally by EKUB. Such schemes are savings and credit co-operatives, whereby the members make regular contributions and then can take advances either in rotation or by competitive bidding. Repayments are cycled back into the scheme so that other members can then take advantage.

For the urban bus fleet of Anbessa, its arrangements for financing fleet acquisition have been described in detail in Section 4.4 of this report.

Fleet inventory

The referenced reports indicate a great deal of uncertainty as to the parc of minibuses and modified taxis operating in Addis Ababa. For registration purposes these had been placed in the same category as sedan taxis, and available figures gave only the combined total.

Previously the Federal road transport authority had kept separate figures for the two vehicle types, but these data were not subdivided by administrative region. In 1997/98, there were 6,049 minibuses and 3,786 sedan taxis in Ethiopia, of which some two thirds were estimated to be operating in or around Addis Ababa. Growth rates for the two categories had been 35% and 84% respectively, indicating both the impact of the 1992 deregulation and the inability of Anbessa to meet inherent demand.

In 2003, the Transport Authority in Addis Ababa introduced a new registration plate design incorporating a barcode for purposes of administrative control. It can reasonably be presumed that only operational vehicles would have applied for the new plates, and so figures from that year forward should reflect the actual vehicle parc in service. 13,290 taxis (of both sizes) were registered in 2003/04, and this figure had increased to 13,697 in the second half of 2004. The rate of annual growth indicated by this last figure is only 6%, and indicates that the market for these vehicles has now saturated.

The Transport Authority is now computerising its records and this will enable a more detailed analysis of its database. A preliminary examination of these records for this report showed that just over 7,500 of the taxis were minibuses or modified taxis, and that just over 5,500 were sedan taxis. This ratio is broadly consistent with the earlier Federal figures if their then higher rate of growth had continued. Observed vehicle condition would also suggest that there is a higher level of owner-drivers who have a greater incentive to keep their vehicle in service and so less sedan taxis may have been retired since 2003.

Regarding the urban bus fleet, Anbessa now owns around 600 buses. However 7 of these are deployed in the remote city of Jimma, and another 70 old buses are unserviceable. As such, its effective fleet is 524 buses, 464 of which are 9 years old or less and are used on scheduled urban services. The balance of 60 buses are 14 years old or more, and are only used for contract services.

Fleet maintenance

None of the taxi owners interviewed for the study on improving urban transport advised that they used preventive maintenance techniques to ensure their vehicle’s condition. Rather the practice was for the driver to arrange (and pay) for any minor repair needed during the course of the day’s operation, and for the owner to take responsibility for major repairs as the need arose and for annual testing.

There is no formal definition of minor repairs, but these are generally taken to cover puncture repairs for tyres, engine oil top up, and minor electrical items such as bulbs. Downtime for the relevant fitment is minimal, and the work would normally be undertaken at the terminal or possibly at a roadside facility.

More major repairs are generally taken to ‘garages’, and these have a good reputation arising from a scheme of government inspection and classification according to their facilities and assessed output. However many owners will also attempt roadside repairs, for which the simple vehicle technology and size of components present few obstacles.

The actual standard of repair is constrained more by the willingness of owners to pay for work that they do not consider to be essential for the continued operation of the vehicle, and the availability of the appropriate spare parts. Until recently the standard of any technical inspection to which the vehicles were submitted was patchy, but this is now being addressed through the registration of 18 private-sector test stations to address capacity limitations at the Transport Authority.

Anbessa has three well-equipped depots and a central workshop at which its vehicles are maintained and repaired. These facilities and practices are described in Sections 4.7 and 5.1 respectively of this report.

Up until now, Anbessa had not been carrying out formal capital rehabilitation on any of its fleet but rather had accepted a gradual deterioration in the technical performance of its buses as they aged. However they have now received a capital grant from the German government of Birr 12million for chassis parts to enable rehabilitation of the best 60 vehicles from its old fleet of Mercedes 1621 buses. This programme will also involve the bus body facility of Abay Technical Enterprises, which has recently been amalgamated with Anbessa.

Vehicle operating costs

Vehicle operating costs for minibuses have been estimated for the preparation of the income and expenditure analysis given in Section 5.4 of this report.

Daily direct operating costs were estimated at Birr 140 for fuel, Birr 60 for road crew, and Birr 5 for marshals. At the assumed utilisation of 130km, these equate to 108, 46 and 4 cents per kilometre (cpk) respectively.

The daily fixed cost for the driver is his hire of the vehicle at Birr 100, but his salary of Birr 20 is paid out of this.

From the owner’s net income of Birr 80 per day, he must pay the variable costs of repairs and tyres. These have been guestimated at Birr 40 per day, or 31cpk.

His fixed costs are income tax (Birr 482 if the vehicle is less than 15 years old) and annual business permit renewal (Birr 60). In general, no third-party insurance cover is taken.

The vehicle operating costs for Anbessa can be taken from the management accounts for 2003/04 given in Appendix 2 to this report. Unfortunately, though, the reporting format set by the public enterprises supervisory authority means that all employment costs are consolidated and not reported separately by functional department.

The government auditors have made a separation of operational employment costs from those related to administration, but have included vehicle maintenance under this heading. In any case, audited accounts are some 4 years in arrears.

The main directly variable costs were fuel (Birr 41.52mn), spare parts (Birr 20.95mn), tyres (Birr 6.49mn), and tickets (Birr 2.95mn). Taking the total level of operation in 2003/04 as being 21.6 million kilometres (though some of these will have been run on contract or in Jimma), these costs equate to 192, 97, 30, and 14cpk respectively.

The semi-variable cost of labour (Birr 31.41mn) equates to 145cpk. The fixed costs of other administration (Birr 7.03mn) and depreciation (Birr 59.06mn) would equate to 33 and 273cpk respectively.

Existing regulatory arrangements and institutions

Review of market conditions

The governing legislation for urban passenger transport is the 1992 Proclamation to Provide for the Regulation of Road Transport, brought in by the Transitional Government after the overthrow of the Dergue. Subsequent legislation allows each Regional administration (including the City of Addis Ababa) to develop its own regulatory framework, but this opportunity has not yet been availed.

In effect, this proclamation deregulated the supply of such transport in line with the prevailing orthodoxy of the time. It was anticipated that new market entrants would be attracted to the business opportunity so created, and this would both improve service quality and reduce the financial burden on the public sector in its provision. However, public administrators had regarded the monopoly franchise held by Anbessa as not having been affected by the proclamation, and so spurned approaches by potential investors.

Any intending operator can readily procure a business permit from the City administration (the Transport Authority acting on behalf of the Bureau of Trade and Industry). The only provisos are that he must have a vehicle with a valid certificate of roadworthiness, and must employ a driver with the appropriate category of licence. The cost of the permit is Birr 110 for initial administration, and Birr 60 for annual renewal.

This permit covers the whole of the conurbation, and is not confined to a specific route or area. Although the 1992 proclamation encouraged the self-regulation of the sector on an area basis, this was not made a condition of the licence. The three owners associations that have since emerged draw their membership from across the city.

The public service vehicle has to be registered as such, either when first imported, when it is converted to passenger carrying configuration, or when on-sold by an existing operator. At that time it must be subjected to a test of its roadworthiness by the Transport Authority, and then re-tested annually thereafter. However, until recently, the effectiveness of the test regime had been restricted by a lack of institutional capacity. This now been addressed by the addition of 18 private testing stations to that of the Authority, though it is too early to say what impact this will have. Examination of the contract between the Authority and stations suggests that there may be insufficient performance incentives, and ineffective sanctions against dereliction of duties.

Driver testing is also a responsibility of the Transport Authority. A Class 3 licence is needed to drive any taxi (sedan or minibus), and a Class 4 licence for larger buses. Testing is usually conducted in an appropriate vehicle, but trucks may be substituted for large buses. Although the testing system is considered to be effective, the security of the resultant licence is not. It takes the form of a sealed pouch inside which the licence details and a photograph are entered onto a card. Apparently it is relatively simple to open the pouch and alter its contents, even changing the photograph. New technology is shortly to be introduced to make this far more difficult, and tenders have already been issued for supply of the appropriate equipment.

It should also be noted that existing legislation only places requirements on the driver and the vehicle providing transport services, and not on the owner. In order to raise standards in the industry, there is a case for the introduction of a specific license to operate a passenger-carrying vehicle for hire or reward. Such a license would be applied to the person keeping or using the vehicle, and not to the driver who is acting as his agent. Specific conditions could be placed on the license holder in respect of maintaining the vehicle in a roadworthy condition appropriate to passenger services.

Fare and fare change mechanisms

Different regimes exist for the setting of fares on the Anbessa buses and the private minibuses and the actual levels of fares authorised vary widely as a result.

Anbessa has an authorised adult fare for each route it operates; this fare then applies for each boarding on that route, irrespective of its individual length. Originally these fares had a direct relationship to the kilometre length of the route, and have officially been fixed since 1992. These rates were Birr 0.25 up to 10km, Birr 0.50 up to 18km, Birr 0.65 up to 23 km, and Birr 1.00 up to 30km. Higher rates applied to peri-urban routes, with a maximum fare of Birr 2.00.

However this relationship has become eroded over time. Whenever a new route is launched, or an existing route is altered or lengthened in response to changing demand in the expanding city, Anbessa adjusts the fare to reflect the costs of the extra operation. The Board then approves this new fare, which in its turn becomes fixed. Further, Anbessa has been allowed to charge a Birr 0.10 premium for limited-stop or express services. Over time, the fares on the routes over which such services were offered have become consolidated at the higher level.

As a result of these various alterations, there is no longer any clear linkage between the length of the route and the fare charged. The Birr 0.25 fare now covers routes up to 8.9km, Birr 0.35 from 6.3 to 10.6km, Birr 0.50 from 8.6 to 14.4km, Birr 0.65 from 11.0 to 12.0km, Birr 0.75 from 17.7 to 19.4km, and Birr 1.25 from 23.5 to 25.8km. Peri-urban routes over 30km have fares from Birr 1.50 to Birr 3.00.

These alterations have resulted in a creep in the fares actually being paid by passengers for their daily journeys. Back in 2000, the Birr 0.25 fare covered 53 routes and the Birr 0.50 fare a further 20; only 7 routes had higher fares. Now in 2005, the Birr 0.25 fare applies only on 15 routes, a Birr 0.35 fare on 12, and Birr 0.50 on 47; 15 routes have higher fares. The typical fare rate paid has risen from 4.0 to 4.6 cents per kilometre for those travelling the full route length. Those making the same journey as before, but now on an altered route, have experienced an even higher increase.

Student fares had also been fixed over the same period, with a rate of Birr 0.15 applying per trip on any route. Clearly this was an increasing anomaly as the Anbessa network expanded, and a new student fare of Birr 0.30 for routes over 9km was introduced in January 2004. By June 2004, sales value of the longer distance tickets had passed those for the shorter routes.

Minibus fares and fare structures have also altered over this period. Back in 2000, Birr 0.50 applied up to 4km, Birr 0.85 up to 8km, and Birr 1.25 for any longer trip. This structure was set so as to discourage their operation of longer routes, with the expectation that Anbessa would then cover these.

The current authorised fares are now Birr 0.60 for 2.5km or less, Birr 1.00 up to 8km, Birr 1.30 up to 11km, Birr 1.45 up to 13km, and Birr 1.75 up to 16km. These new tariffs have been set after negotiations with the owner associations, and have been designed to help them recover the increasing cost of fuel in recent years. Further increases are currently under consideration, probably amounting to Birr 0.05 on the shortest routes, and Birr 0.10 on all others.

However the revision in the fare structure, whereby the second level of fare is now applicable to routes over 2.5km, and not 4.0km and above as formerly, has created anomalies with regard to the fare rate paid. All journeys up to 4.2km achieve a fare rate of 24cpk or over. Longer routes, though, show a sharp decline in their revenue potential. Routes from 6 to 10km earn roughly 15cpk, and this falls to 13cpk for those that are even longer.

Whilst this fare structure still discourages longer distance operation, it has now created an opportunity for fares manipulation. If a trip of, say, 6km can artificially be broken at an intervening point, then separate fares of Birr 0.60 and 1.00 can legitimately be charged. The resultant fare rate has been increased to 27cpk, consistent with the earnings on shorter routes. However the passenger has suffered inconvenience, and additional congestion has been caused at the interchange station.

Analysis of the income and expenditure for minibuses suggests that such higher fare rates are actually needed for their financial survival. In order to provide this, a new fare band for trips above 5km needs to be introduced and the rate set appropriately. This alteration would have a greater benefit for the system than the simple fare increase being proposed.

Effects of unions and union regulations

Market entry is not controlled by the owners associations, which are relatively ineffective in the present regulatory framework, or by the marshals unions. However the marshals unions do affect the orderly working of the market, through controls on queuing at the stations, and impose charges on the minibus drivers accordingly. This has removed the need for any direct competition between drivers at the stations, and the conflict and congestion problems in which this would result.

However, in order to preserve the net earnings of the drivers, the marshals impose a number of operating practices that work against passenger interests. The principal of these is waiting for the assembly of a full load before commencing operation on the route. This can impose high waiting times at terminals, particularly outside peak times, and makes it very difficult for intending passengers to access the service along the route near to the terminal. The latter, in turn, then requires an extended walking leg from the point of journey origin to the route terminal to ensure access. It is also normal practice to force the waiting passengers to sit in the vehicle in the full sun, rather than to access any available shade.

By loading vehicles in strict rotation, the marshals prevent intending passengers from rejecting ones that fail to meet expected standards of cleanliness or physical condition. This, in turn, lowers the incentive for vehicle owners to improve their performance in this domain in the absence of any effective external enforcement. Any investment in a premium quality vehicle also becomes impractical under these circumstances, and acts as a barrier to the development of a service offer capable of attracting existing car users.

Taxation and other incentives / constraints

Minibus owners and drivers are assessed to income tax on the basis of presumed earnings, rather than any explicit declaration. Whilst this is a prudent measure to broaden the tax net, it may result in an under collection if the assumptions used prove not to be realistic. From the analysis made in Section 5.4 of this report, this may well prove to be the case. Further, the rates of income tax charged are progressive – that is, higher rates are assessed against higher incomes. If multiple ownership exists, a standardised collection for each vehicle will again result in an under collection.

Taxes and duties on operational inputs are broadly consistent with other business sectors, and do not create any specific incentives. However the lower excise duty on higher capacity buses (10% against 35% for those with less than 16 seats) is a useful incentive for operational efficiency.

Anbessa is a special case in respect of incentives, receiving both operational and capital subsidies. The City of Addis Ababa makes a payment to the enterprise to compensate for the public service obligation imposed by sub-economic fares. The Federal government, the owner of the enterprise, has provided direct capital grants and has directed external donor assistance in respect of both grants and concessionary finance.

Operational subsidies were introduced in 1996/97 and equated to Birr 0.24 per passenger carried. However it was always intended that these subsidies would be transitional during the revival of the enterprise led by the Dutch technical assistance programme of that time. These had been progressively reduced in recent years, and had reached Birr 0.11 per passenger in 2004.

However, in the latest grant from the City, the subsidy has now taken the form of a fixed sum of Birr 20 million, expected to equate to some Birr 0.08 per passenger. The removal of the linkage to passenger carryings takes away any direct performance incentive from Anbessa, and also suggests that the subsidy is to the enterprise rather than its customers. Whilst this may act to simplify budgeting at the City administration, it represents a regrettable precedent.

Perceived problems

Inadequate service quantity

An assessment of the adequacy of service quantity can only come through indirect indicators. From the passenger perspective, inadequate supply will manifest itself either in excessive waiting times for accessible services or in unacceptable levels of over-crowding on the vehicle. From the operator perspective, inadequate supply would normally result in higher prices and excess profitability. However, where fares are subject to over-zealous control, these signals can be masked and low profitability can co-exist with unsatisfied demand.

The study on improving urban transport found that 90% of minibus riders indicated that the long waiting time for a bus service was one of the reasons for their modal selection. For the bus users themselves, the frequency of service (and hence resultant waiting time) was rated as bad by 60% of those interviewed. Unfortunately, though, average waiting times for the two modes were not surveyed, or variances identified in respect of peak and off-peak times for example.

According to the same survey, over-crowding on the buses was not an important factor in their modal choice for minibus users. However for bus users crowding and overloading was the worst rated quality of service indicator, rated as bad by 75% of those interviewed. This concern can be supported from Anbessa’s own route performance data that show the average number of adult passengers per trip at 104. Clearly achievement of such levels has to indicate peak load factors as high as 150% on occasion.

Taken together, these indicators clearly demonstrate a shortage of supply for conventional low-cost bus services in comparison with the level of demand for these. Part of the reason for the excess of demand must come from the fares control that has resulted in a significant decrease in the real level of bus tariffs over the 12 years that they have effectively been frozen. However passenger affordability is a key concern of the City administration, and it is unlikely that this policy will see a near-term reversal given that fares increases were not allowed in the face of the escalation in fuel prices in the last few years.

The picture for minibuses is far less clear. The study on improving urban transport suggested that these were in gross oversupply, based on observations of the taxi stations in the inter-peak periods. However this is more of a reflection of their lack of competitive pricing in comparison with bus services for the discretionary travel typically made at these times, and of the higher service quality provided by the buses outside of the peak hours. On almost all bus routes, there is a higher effective frequency of service and a lower level of over-crowding in the middle of the day.

Observation at selected stations in the peak hours indicated a clear variance between the so-called ‘hot’ routes and the rest of the minibus network. For the former, it was more normal for passengers to be queuing for vehicles; for the latter, vehicles were queuing for passengers. The explanation appears to lie in the hilly terrain, especially in the north of the city, which requires a powerful vehicle in good condition to sustain the operating conditions. These are in relatively short supply, but the resultant passenger waiting times did not appear excessive. Unfortunately, though, these have not been surveyed.

One indicator that transport supply and passenger demand may be coming into line even on the ‘hot’ routes is anecdotal evidence relating to the value of the vehicles used for these. A few years ago such vehicles could command a price of Birr 120,000 to 130,000 (ca. $15,000); today the price is reported to have fallen to about Birr 70,000 to 80,000 (ca. $9,000). This lower level would be more typical for similar vehicle values in neighbouring capitals, such as Kampala and Nairobi, and would tend to indicate that excess profits are no longer being made.

For the older and less powerful minibuses, and particularly the modified taxis, it is almost certain that supply now exceeds demand. One indicator is that the number of registered taxis declined when the new plates were introduced in 2003, suggesting that those only making a marginal return or in need of extensive repair were withdrawn from service. The physical condition of many of the remaining vehicles still leaves much to be desired, though, and a higher standard of roadworthiness inspection (resulting from private participation) will act to reduce numbers further.

Another key factor in the supply / demand balance will have been the considerable expansion in Anbessa’s operating fleet in recent years, particularly the 150 new buses acquired in 2003. The older minibuses, and particularly the modified taxis, no longer provide the higher quality of service that can command the tariff premium being charged. These vehicles are therefore being marginalised, and often confined to market-related activities where the carriage of goods and produce provides a significant proportion of income.

Low levels of safety

Surveys conducted for the study on improving transport did not identify safety as being a key concern for either bus or minibus passengers. This view is probably supported by statistics from Anbessa that indicate that their accidents, whilst being quite frequent, are very rarely serious. Urban bus passengers are normally only exposed to danger when attempting to board or alight from a moving vehicle, or when vehicles are being manoeuvred at a terminal. The high operating standards of the enterprise make these conditions relatively rare.

The Traffic Police in Addis Ababa register all reported accidents, and assign these to the type of vehicle involved; unfortunately, though, this assignment does not recognise the severity of the accident in human or material terms. In 2003 there were some 2,400 accidents involving taxis, out of a total of 10,200 for all vehicle types. Both figures represented an escalation over previous years, but the proportion of accidents involving taxis has stayed relatively constant at just over 20%.

Whilst taxis only represent about 11% of the vehicle parc, their higher than average level of utilisation suggests that their safety record is no worse than for general traffic. In fact, with just over 13,000 taxis being registered, the accident rate works out at less than one for every five years of operation. Clearly this indicates that minor accidents are not being captured by the official statistics.

The real safety issues with regard to urban transport in fact relate to pedestrians. Very little provision is made for their safe mobility within the road infrastructure and, where sidewalks are provided, all too often these are commandeered by hawkers or parked cars obstructing people movements and forcing them into the roadway. Road crossings are also problematic, with few signalised junctions having adequate pedestrian provision, and unsignalled crossings mostly being ignored by motorists; there are very few pedestrian over-bridges or underpasses in the city. Further problems exist at public transport stops and terminals, with inadequate segregation of movements and little consideration being given by drivers.

Poor service quality

Surveys conducted for the study on improving transport found that quality of service was a significant concern of bus users, but was not identified as an issue for minibus riders. 75% of bus passengers rated crowding / overloading as bad, 40% were concerned about personal safety and petty crime, and 33% found service reliability and breakdowns unacceptable.

These results are not surprising given the high load factors identified on the bus services. The resulting conditions are not only uncomfortable in themselves, but also foster an environment where petty crime can proliferate both on the vehicle and in the boarding queues. The fact that minibuses are rarely overloaded (at least in the city centre) and that queuing times appear to be shorter, may explain why similar concerns have not been captured.

Such issues are likely to affect women more seriously than men, but passenger responses in the survey were not stratified by gender. However the fact that women only formed 28% of bus passengers interviewed, as against 43% of minibus passengers, suggests that this is in fact a significant factor in modal choice. When the relative disposable income of men and women is also taken into account, the apparent impact is even more striking.

It is unclear, though, whether the concern about bus service reliability is more a reflection of the incident rate or of the consequences of a breakdown. Clearly the affected passengers will find it extremely difficult to board another vehicle, given its likely level of occupancy, and a direct service replacement will take some time to organise. A similar effect should apply for the minibuses, but the numbers involved in each incident will more readily be absorbed into the traffic flow.

Both the study on improving transport and the urban transport study comment on the lack of facilities provided for passengers, both at the terminals and at stops along the route. Even the main terminals lack adequate toilet facilities, and these are not provided at the secondary interchanges. Only a very small proportion of the 900 bus stops have any form of shelter, and these have not been incorporated into even the recently constructed interchanges.

Passenger information is also lacking throughout the system. Most bus stops do at least show the route numbers being served, but there are no timetables or indication of destination for casual passengers. Information standards are better at the terminals and interchanges, but these seemed to attract little interest from passengers. Either passengers already know the system well, or the information provided has proven not to be credible.

Problems with the vehicles centred on their general cleanliness, the condition of their seats, water leaks during the rains, and the attitude of the crew. Focus group discussions identified cleanliness as a very real issue when the vehicle was also used for the carriage of produce or market goods, as sometimes happened in periods of low passenger demand. However the general standards in the sector are not acceptable, based both on the typical age of the vehicle and a repair culture that tends to focus on its mechanical aspects rather than the passenger saloon.

Concerning the physical condition of the vehicles, this reflects both the maintenance culture in the industry and the lack of effective enforcement capacity at the government level. As the industry is effectively self-regulated, the owner associations could also do more to raise the standards expected of their members. However their funding through charges levied on these gives them no incentive to expel those who fail to perform. Further, the operating system of strict rotation of work at the terminal does not allow passengers to select vehicles based on their condition.

With regard to the standards expected from road crew, these have no strong basis in law and lack effective enforcement. The driver of a commercial passenger vehicle of any size is required to have a degree of maturity and experience reflected in a Class 3 licence. However the Transport Authority reports that existing licences can easily be tampered with, and hence that possession of the appropriate licence is not a necessary indicator of capability. An action programme has now been devised to tackle this problem, and tenders have already been issued for the supply of the appropriate equipment.

Minibuses carry an ‘assistant’ to collect fares and manage passengers, but no conditions are set down with regard to the wearing of uniform or other relevant standards of behaviour. In most cases the driver employs the assistant on a daily basis, with the main criterion for his selection only being his honesty in declaring all fares paid.

Low level of affordability

Affordability of urban bus fares has been a primary concern of both the Federal government and of the City administration. As a consequence, approved fares have been held at 1992 levels despite the significant increase in the cost of all operator inputs over that period. The relative affordability of fares has therefore improved greatly.

The study on improving urban transport identifies the average trip cost for bus passengers as being Birr 0.80, but this seems unlikely as it would imply a travel distance in excess of 15 kilometres. At the weighted average travel distance surveyed of 7.2 kilometres the typical fare would be Birr 0.35, with anything from Birr 0.25 to Birr 0.50 being possible. A regular bus commuter would therefore spend some Birr 15 per month.

The same study identified a weighted average personal income of Birr 450 per month, which is not implausible for those in regular employment. For such people, the cost of travel works out at only just over 3% of income, a very low level by comparable international standards. Even if Birr 0.80 were being paid, the percentage of income spent on transport would still be less than 8%.

For minibus users, the study on improving urban transport found that the average trip cost was Birr 1.7 Birr. This reflects the high proportion of such trips made at peak hours when the operators can exploit the fare structure by breaking journeys into smaller legs and hence the effective fare rate. Monthly costs for a commuter would then be some Birr 75, representing around 11% of their higher weighted average personal income of Birr 690. This ratio would still not be considered high by international standards, but reflects the relatively short trip lengths arising from the compact urban form of Addis Ababa. The weighted average trip length surveyed for minibus users was only 5.5 kilometres.

The urban transport study has conducted a large number of household expenditure surveys, from which it should prove possible to verify these finding, but its data are not yet available. It would be hoped that this study will also identify the spatial distribution of incomes, as it is not uncommon for low-income households to be forced to live at some distance from the city by rents.

Were this to prove to be the case, then higher travel costs would coincide with lower incomes and affordability would be a more significant issue under these circumstances. The study on improving urban transport found that 60% of pedestrians walking more than two kilometres were not able to afford using public transport services. Further, 83% of bus passengers were not able to afford minibus fares. Clearly there must be a high level of income disparity, and affordability is a very real issue for at least parts of the urban population.

Low operating speeds

The study on improving transport found that the shorter travel time available on minibuses was a factor in their modal selection for 73% of passengers interviewed. Even though this proportion was lower than the 90% of passengers who were influenced by waiting times, it does indicate that operating speed was still a significant issue.

Even 28% of bus passengers stated that their modal choice was based on faster travel times, though in this case their opinion is likely to have been based on minibus operating practices designed to maximise their load factors. These include calling at all possible boarding points to tout for business, and loitering for a replacement passenger when one has alighted.

No figures are available for the operating speed of minibuses, but the scheduled speed on Anbessa services averages 16.8 kilometres per hour across its network. Whilst this figure includes an allowance for trip recovery time and static ticket sales at terminals, the high level of lost trips (15%) suggests that this figure is overly optimistic particularly at peak times. Nevertheless such speeds are relatively high for an urban bus system with no public transport priority measures, and reflect the high physical capacity of the primary road network

As well as operating speed being a high direct concern of minibus passengers, it also has a strong indirect impact on the price that they must pay for their service. Low operating speeds lead to low productivity for operators, reducing their gross revenues and recovery of their fixed costs related to the vehicle and of their semi-variable costs related to the road crew. Further, traffic congestion increases both fuel consumption and wear on the vehicle driveline. Both effects impact negatively on profitability and so lead to the need for higher tariffs.

The urban transport study will identify unacceptable levels of service at peak hours both on sections of the main arterial road network and of sub-arterials in inner-city suburbs; it will also confirm the inefficient use of road space by low capacity modes of transport. The study on improving transport identified that cars moved only 15% of people, while accounting for 60% of vehicles in traffic. However even the minibuses required 23% of all vehicle trips to carry 39% of people, whereas the buses carried 45% with only 4.5% of the traffic flow.

Clearly, under these circumstances there is a need for some degree of priority in the allocation of road space to high-capacity public transport. The urban transport study will make detailed recommendations for the implementation of such measures, which are also addressed in the study on improving transport.

For roads where public transport priority measures are not a feasible option, there is a need to increase their functionality and to introduce intersection controls so as to raise capacity and increase operating speeds. A hierarchy of roads also needs to be introduced, with main arterial routes being segregated from unrestricted local access.

In general, little provision is made in the current road designs for either passenger-transport lay-byes or for queuing space for turning traffic; resultant delays impact on all vehicles. Also relatively few intersections are signalised, and these tend to suffer from inefficient staging and unreliable power supply. Manual control of traffic at intersections is patchy, with Police resources clearly being stretched. The traffic controllers appear to have been well-trained, but they have no means to co-ordinate their interventions effectively.

Poor service accessibility

Unfortunately the primary indicator of service accessibility, the walking time to the nearest boarding point for motorised transport, has not yet been surveyed. However the study on improving transport found that the shorter walking distance to a stop was a factor in their modal selection for 60% of minibus passengers, and that 90% of passengers indicated that there was no available bus service to their destination.

The operation by Anbessa of 80 bus routes to serve the city of Addis Ababa (excluding the 9 peri-urban routes) suggests quite a high level of accessibility of services across the network. However the peculiarity of the fare system means that separate routes are provided in each primary corridor so as to exploit the opportunity to maximise charges. This degree of route overlap results in a lower degree of spatial coverage than the bare figures would suggest.

Further, the need for Anbessa to maximise its load factors (and its underlying lack of fleet capacity) mean that low-density areas are generally not that well served. Routes are only cautiously extended into new developments, and then the justification is more likely to be the opportunity to charge a higher fare over the whole route than to serve an unproven local demand.

Finally, neither of the two referenced studies has identified any significant use of either non-motorised or intermediate forms of transport to enable easier access to the Anbessa network. The hilly terrain of much of Addis Ababa probably explains the lack of bicycle taxis that are often observed in other cities of the region, but there would appear to be no obvious reason for not using motorcycles. However there are reports of animal transport in outer suburban areas, and this may be more significant for access to the peri-urban services.

The minibus network provides a higher density of coverage, with 106 routes serving the city and a lesser degree of overlap than on the Anbessa network. The modified taxis, in particular, are able to offer services on unmade roads that would not be considered suitable for Anbessa. For many users, service accessibility is more important than price and suggests a willingness to pay for a better quality of service than that provided by Anbessa.

Inappropriate vehicle type and size

Public transport in Addis Ababa is provided by high-capacity standee buses, with a nominal capacity of 100 passengers, or by minibus and modified taxis with a capacity restricted to only 11 passengers. Shared sedan taxis are a very low proportion of the modal mix, and there are no urban rail services at present.

The large urban buses are clearly an appropriate response to the levels of passenger demand in the main corridors, but less so when individual routes are operated at headways exceeding 20 minutes at peak hours. This anomaly arises from the peculiarities of the fare system such that Anbessa has an incentive to differentiate its routes so as to maximise the rates that it can charge on each service. A revision in the fare structure to an explicitly graduated or zonal system would allow an integration of services, and generally much shorter headways.

Where the passenger configuration of these large buses is less appropriate is on the peri-urban routes that can extend to as much as 47 kilometres in length. Under these circumstances, it is clearly more comfortable and safer to provide seats for as many passengers as possible. With a single-door body it should prove possible to provide seats for up to 65 passengers, and some regional regulations would allow a further 25% of standees resulting in an overall capacity of 81. This number would not be much different to the average loads being achieved at present, and so need not result in significant fares increases.

The fact that there are no urban buses offering a passenger capacity between 100 and 11 is a direct consequence of two regulatory decisions. Anbessa was granted an exclusive franchise for the provision of urban bus services, but the tight control over the fares that it can charge required the lowest possible cost per passenger kilometre and hence the highest capacity vehicle that was operationally suitable. The only other vehicles allowed to operate urban passenger services were taxis, and these are restricted to the carriage of 12 people including the driver.

Deregulation of the passenger transport industry in 1992 might have been expected to result in the emergence of urban bus operators other than Anbessa, but these were not encouraged. The tight fares controls, and the access to both capital and operational subsidies given to Anbessa, have provided increasing barriers to competitive market entry in the intervening years. However there appears to be no reason why commercial bus services, operating at a fare premium to the subsidised services so as not to be extractive, should not now be solicited.

At the other end of the supply spectrum, the restriction on taxi passenger capacity was set originally to ensure the safe operation of the ‘modified’ taxis that were converted pick-ups. Given the relatively long rear overhang on these vehicles, any greater number would result in a dangerous reduction of the load carried by the front axle resulting in the loss of steering control and ineffective braking. Such an effect was readily identifiable for similar vehicles in Nairobi where no such controls were put in place. In any case 12 people would weigh at least 0.8 tonnes, which would be at the safe capacity limit once the weight of the bodywork had been taken into account.

When the importation of vehicles more appropriate to passenger services became possible, such as the Toyota HiAce and the Nissan Urvan, the loading restriction remained in place. In neighbouring countries, such as Kenya and Uganda, such vehicles are allowed to carry up to 14 passengers using the fold-down seats for which they had been designed. In some countries, up to 19 passengers have been permitted, but this does result in a high degree of passenger crush.

Any relaxation in the authorised capacity for these vehicles would enable a reduction in their breakeven fare, and a decrease in the number on the road needed to move the same number of people. The result would not affect Anbessa in any significant manner, as the fare differential would still be high enough to discourage most bus passengers from transferring to this mode.

Were the limit to be raised further, say to the 25-passenger threshold that applies to matatus in Kenya or tro-tros in Ghana, then the importation of higher capacity commercial vehicles for conversion to passenger use would become practicable. Duty rates for buses with more than 15 seats are lower than for smaller vehicles, and this would further assist in reducing their breakeven fare. However this effect might then begin to become extractive from Anbessa’s services, and economic regulation would be required to avoid wasteful competition.

On the other hand, the very large number of routes offered by the industry requires that small buses often be used to provide an attractive frequency particularly in the off-peak and in the peripheral areas. This issue will be examined further in the later Section on efficient network planning. Clearly a stratification of vehicle capacity to match passenger demand on each route would provide a more efficient outcome.

Much of the urban transport fleet in Addis Ababa offers low standards of passenger accessibility and comfort irrespective of the typical vehicle condition discussed in the following Section.

As noted earlier, Anbessa buses are operated in a standee configuration that provides only 30 seats. Further, these are conventional truck-derived vehicles with high floors and entry steps that result in difficult access for the encumbered or mobility impaired. Even the modest access standards set in UN ECE Regulation 36 are not met, especially when the bus is first being boarded at the terminal.

The modified taxis also provide difficult access to their rear cabin, with a high first step and then a restricted roof height once inside. The lateral bench seats provide inadequate support for passengers under braking or sharp cornering. It is also reported that their body design makes poor provision for ventilation, and this is a contributory factor to their retention of odours. By contrast, the minibuses offer easy access and movement through the saloon though their ceiling height is still low.

It is also noticeable in Addis Ababa that no attempt has yet been made to provide a high-quality public transport service other than the minibus concept. If diversion from car use (or restraint on its growth) is to be obtained, then a service offer acceptable to car users will need to be made available to complement car demand management measures. This will not come from the so-called premium buses suggested in the study on improving urban transport, as these would retain essential bus characteristics.

What is needed is a high-quality midi-bus service of the kind developed in Nairobi by Kenya Bus Services Ltd, as reported in SSATP Working Paper No.75. Clearly the development of such a service cannot be mandated, but at the very least the regulatory framework must place no obstacles against its possible implementation. This would apply in particular to any control on the ceiling for fares to be charged, though a floor might be set at (at least) 50% more than those for conventional services to avoid extraction from the core network. Public investment might be provided for infrastructure needed to support such services, such as park-and-ride sites and passenger shelters in central commercial districts.

Poor vehicle condition

As has been noted earlier, the physical condition of vehicles operating in the urban passenger transport industry in Addis Ababa often leaves much to be desired. However this does not only concern their passenger saloons, but also their general mechanical condition. Casual observation of the vehicle parc shows many vehicles with damaged lights and smooth tyres, as well as some with dense exhaust smoke.

In Ethiopia there is no commercial vehicle operator licensing system for taxis and minibuses, with only the vehicle and its driver being subjected to examination against set standards. Thus there are no general standards set for operating performance in the passenger transport industry, and (critically) no explicit duty to maintain a vehicle in a roadworthy condition. Further, there has been no explicit requirement to insure against third-party and passenger risks that would provide an incentive for underwriters to take an active role in ensuring vehicle condition and hence reducing their risk.

Without the incentive provided by effective detection of vehicles in a dangerous mechanical condition, or effective sanctions against their continued operation, it is most unlikely that the prevailing maintenance culture in the road transport industry will change. Vehicle operators are reported as using corrective repair rather than preventive maintenance in ensuring their vehicle condition. The costs of repair were seen as being high, caused by the high price of spare parts, but the rates paid for labour were reasonable and the standard of work reported as satisfactory.

Unfortunately, the informal operator working within a constrained cashflow sees preventive maintenance as an avoidable expense and lacks the information to make rational choices to minimise his vehicle’s life-cycle cost. The end result is the poor standard of vehicle condition widely observed, even though in-service reliability seems surprisingly high. Some advantage must come from the regular hiring of vehicles to specific drivers, resulting in incentives for mechanical sensitivity as well as learning to drive round the faults of the vehicle in question.

On the positive side, technical skills in the private garage trade are reported as being good, and the actions taken by the Federal road transport authority to classify garages according to their facilities and competence has helped to raise standards. Thus, when repairs are actually undertaken, the work will generally be well done if the relevant spare parts are available.

All commercial vehicles are required to submit to a technical inspection on first registration and then yearly thereafter. However the Transport Authority had lacked the needed institutional capacity to carry this out in an effective manner. Until recently, it was only able to offer a single testing station and capacity constraints meant that inspections were often cursory even when they were carried out at all. Further the technical facilities available to the Authority were also inadequate, particularly in respect of larger commercial vehicles. Special test equipment was also lacking, especially with regard to exhaust emissions.

Now, though, it has brought 18 private testing stations into the network though it retains technical control over the setting of standards. It is too early to say how effective this measure will prove, and the Authority still needs to put appropriate monitoring and disciplinary procedures in place. Potentially, deterrent penalties could be put in place that would prevent the private stations or their employees risking providing certification to any unroadworthy vehicles.

A further problem with regard to vehicle quality is the average age of those operating in urban passenger service. As was noted earlier the typical age of modified taxis is at least 15 years, and this is explicitly recognised in their assessed income from transport operations and hence the quantum of tax levied on this. Not only will vehicles of this age have suffered abnormal cumulative wear, but it also becomes increasingly difficult to source their spare parts. Thus, even if a full repair is desired, it may only prove possible to complete part of the necessary work.

While the minibuses being used on the ‘hot’ routes will be somewhat younger, others may be nearly as old as the modified taxis. Similar problems will exist in keeping these in roadworthy condition, but at least there is a high degree of interchangeability of components with more modern vehicles and a parallel market for their spare parts has emerged.

The business permits issued by the Transport Authority on behalf of the Bureau of Trade and Industry are not assessed against network needs, and are granted automatically if the supporting documents are in order. Although the Transport Authority intends to introduce regulatory reforms, these are based on enhanced self-regulation by the owner associations rather than the introduction of competition for the market under its own direct control.

Inefficient operating procedures

Control of the minibus sector by the ‘marshals’ results in a number of operating procedures designed to maximise their own returns rather than prioritise customer needs.

All routes both start and finish at terminals controlled by the marshals, and longer routes will also pass through their interchange stations. While the general location of these terminals in the suburbs is relatively close to the point of demand generation, those in the city centre are often at some distance from the point of trip attraction. Partly this reflects the availability of suitable land in the CBD area.

A further problem arising from the use of CBD terminals is the congestion that these attract through the intensity of vehicle movements, the density of pedestrian flows to and from the terminal, and the generation of informal economic activity (‘hawkers’) taking advantage of the customer base so created. Finally, a route network based on a static terminal structure is unlikely to remain responsive to changes in passenger travel patterns. It can also affect the number of service interchanges that passengers have to make within the network in order to complete their journeys.

This situation is made worse in Addis Ababa by the authorised fare structure that results in a sharp drop in fare rates for trips over 4 kilometres. Under these circumstances, it makes sense for the driver to break his route artificially on a longer trip so as to be able to charge two fares for the shorter legs at the higher rates applicable to each. Clearly this is in the interest neither of the passenger nor of the transport system.

The typical pattern of operation in the industry is for the vehicle first to be fully loaded at the terminal before being despatched on the route (‘fill and run’). As has been noted earlier, this does not necessarily result in overly long passenger waiting times at the terminal provided that the overall frequency of service is high. Slightly extended waiting times at peak hours were attributed to the temporary shortage of vehicles on some ‘hot’ routes then rather than to slow loading.

However anecdotal evidence, and feedback from the focus group discussions, indicates that there can be far more severe problems – particularly in the peripheral areas, or on routes with a relatively low level of demand. The surveys also suggested a problem at interchange points, where service in the contra-peak direction was often delayed while operators focused on meeting demand in the peak direction. It was also noted that some drivers will loiter along the route once a passenger has been dropped in the hope of filling that space; this practice is more common in the off-peak.

‘Fill and run’ would appear to be a rational operating practice for an industry that is typified by low fixed and high variable costs. Unless additional vehicle supply on a route were to generate an additional level of demand, it is better to maximise the load factor (and hence revenue) on each trip rather than to maximise the number of trips. Similarly, if there is no immediate prospect for work at the destination terminal it makes sense to maximise loadings on the trip being run rather than to rush to join the next queue.

Given the relatively short waiting time that is being experienced by passengers along the route, it seems unlikely that an increase in frequency would actually attract extra passengers to those surveyed. However the level of service on low demand routes may actually be a deterrent to travel, especially in the off-peak. Under these circumstances it would make sense to offer a maximum headway between services, with these then being operated against a clock-face timetable.

One consequence of the ‘fill and run’ practice is that it becomes virtually impossible to board a vehicle along the route close to the terminal. As such, passengers are forced to walk to the terminal even when another boarding point might be more convenient for them. Regular passengers get to know not to wait at points from which they are unlikely to be able to get a place.

The positive aspect, from the operators’ perspective, is the high load factors that can then be attained as a result. The study on improving urban transport identified an average factor for minibuses of 90% in the peak and 83% in the off-peak. Clearly these figures are at the practical limit unless a degree of overloading at peak times is to be tolerated.

Within the terminals, the marshals enforce a strict rotation of work between the drivers, with the first vehicle in the queue being loaded before passengers are allocated to the next. As has been noted earlier, this leaves the customers with no choice as to the vehicle they might wish to use unless they are prepared to wait and submit to harassment. As a result there is no real incentive for individual owners to improve the cleanliness or attractiveness of their vehicle, as they will receive custom regardless.

Inefficient network design

The proliferation of routes within the Anbessa network has arisen through an unintended consequence of price controls. The fare charged for any journey is the fare applicable to that route, not to the distance of travel by the passenger. Authorised fares can only be raised when a route is redesigned or extended, or a new route is created. This system has provided an incentive for Anbessa to construct a complex network with a high degree of overlap between individual routes. Each of these routes is then run at a relatively frequency of service, and fares on the common sections may vary significantly.

Similarly for the minibus sector, route fragmentation is an unintended consequence of a poorly designed fares structure. When unrealistically low limits were set on fare rates for longer trips, apparently to encourage a modal shift to buses, operators responded by breaking up routes into shorter sections so that they could still earn a return. This practice is particularly prevalent at peak hours when the minibuses are price setters, though it is much less common in the off peak when they are price takers.

If the potential productivity gains offered by larger buses are to be fully realised, and hence reductions in cost per passenger place kilometre be obtained, then the priority must be to redesign the minibus route network based on the principles of aggregation of demand and the use of planned interchange. The total number of routes offered would reduce significantly, but the range of mobility options would be preserved. With a smaller number of routes, larger buses would still be able to offer an attractive frequency of service over these.

The implication of forcing interchange on passengers through a redesign of the route network could be seen as negative in terms of the additional costs imposed on them. However it needs to be recognised that a high proportion of passengers are already forced to make interchanges during their journeys. The study on improving transport found that the 38% of minibus passengers needing to interchange were, on average, making two transfers per trip. Somewhat implausibly, the 57% of bus passengers needing to interchange were reported as making three interchanges per trip. What is important is to mitigate the impact of interchanges on those having to make them through a range of supporting actions.

The most important of these requires minimising the financial cost of the interchange. In the absence of any centralised fare collection system, or acceptance of through-tickets between operators, this will require that the fare structures broadly reflect the distance travelled on the trip. This will need a change in the current practice where a flat fare is charged on each bus route, irrespective of the length of the individual boarding. Similarly for the minibuses a new fare structure will be required that provides for an enhanced fare for trips over 5 kilometres. Clearly there are potential implications for the revenues earned byAnbessa, and a careful balance will thus need to be established between producer and consumer interests.

Given the high concerns expressed by passengers in respect of comfort and security, attention will also need to be paid to the facilities provided at the interchange points and mechanisms be designed for their maintenance in good condition. Initiatives that have been suggested for the development of the central terminals could be applied equally to the main interchanges. At other points in the network where routes intersect, careful consideration will need to be given to positioning of stops and the provision of walkways between them.

Finally, the potential time penalty of interchange also needs to be considered. Fortunately the observed value of time for public transport passengers is relatively low. Nevertheless time delays can be minimised by ensuring a relatively high frequency of service on routes serving the interchange. This will require that a minimum level of service be specified on all routes, if necessary breaking the principle of ‘fill-and-run’ for minibuses at times of lower demand.

Attributed causes

Poor operating standards and inadequate enforcement

Casual observation indicates a relatively low level of driving standards in the road passenger transport industry, and the urban transport planning study identified that driver indiscipline was a main problem. Driver behaviour, especially in the vicinity of passenger pick-up points and around interchanges, has impacts not only within the sector but also affecting general traffic congestion. Competition between vehicles operating on common routes provides incentives for blocking at the stops and / or aggressive driving in order to reach these first. Vehicles stop in the roadway to allow waiting passengers to board, and sometimes even while a load is being assembled. It is quite common for such behaviour to block the two nearside lanes of a dual-carriageway.

Driving behaviour in Addis Ababa is governed by the Road Traffic Safety Regulations (No.5 of 1998), which sets out a range of offences, and the penalties attached to their commission. It also sets out the powers of Traffic Controllers, who are defined as Police officers in uniform, and articulates the hierarchy of responsibilities within the City departments. It is intended that the implementation of these regulations be assisted by the establishment of a Road Safety Council (through Proclamation No.7 of 2003), but this has yet to take effect because of some logistical impediments.

Whilst most of the offences are capable of unambiguous interpretation, those in relation to driving under the influence of alcohol or drugs are not. There is no definition of being drunk in terms of body alcohol (blood or breath) or behavioural response. The taking of drug or chat has no time specified in relation to the offence, and even the term ‘under the influence of’ would have a tighter definition. Admittedly the objective definition of acceptable levels of drug residuals is an international problem, and appropriate technology for their enforcement might prove expensive in the Ethiopian context.

The basic penalties for a first offence against the regulations take the form of a cash fine that rises in step with its severity from Birr 40 to Birr 140. Whilst such penalties might appear to be well correlated to average incomes, they are extremely low in the context of commercial transport. The driver of a minibus might expect to collect daily fares in excess of Birr 300, and his own daily income (salary, expenses, and surplus) could exceed Birr 60. In the case of a private car, the value displayed in its purchase is far in excess of the penalty. As such, these fines cannot be considered as deterrent.

What is encouraging in the regulations is the additional punishment based on the frequency and gravity of the offence. In this context, gravity refers to the extent of any personal injury arising from the incident (though this is not defined in the regulations). Some commentators feel that gravity should relate to the potential of the behaviour to cause injury, and not just to the chance that it did.

Repetition of offences by a driver results in the loss of his licence for escalating periods up to life. However the effectiveness of this sanction does depend on the security of the driving licence, and the Transport Authority acknowledges that this has been a problem. It has now identified an improved technology, and issued tenders for the procurement of the appropriate equipment.

Traffic controllers do, though, have an effective sanction through putting a vehicle under arrest, and seizing its rear number plate if it then has to be removed so that a repair can then be undertaken. However the offences for which these sanctions can be applied relate only to vehicle lighting (admittedly a significant problem), the lack of a roadworthiness certificate or necessary driving licence, and being drunk or having taken drugs.

A vehicle found to be unroadworthy in any of the primary safety areas – tyres, brakes, steering, as well as lights – should be subjected to an immediate prohibition as an annual inspection is not sufficient to protect these standards. Also, the ability of an intoxicated driver to find a replacement to enable the continued operation of the vehicle considerably dilutes the deterrent effect of the penalty in any commercial activity. The immediate lifting of the arrest under these circumstances should be removed from the regulations.

Lack of profitability and investment

Whilst, clearly, it is not possible to glean accurate data on the profitability of the passenger transport sector in Addis Ababa from its dispersed private-sector operator base, the analysis conducted in Section 5.4 of this report strongly suggests that this is now poor. Further, the need for drivers to exploit the current fare structure by artificially breaking passenger journeys also suggests that margins are low.

Profitability can also be estimated from the number of new entrants attracted to the industry and the nature of investment in its rolling stock. In 2000, the annual rate of growth in minibus registrations at the Federal level was reported as being 35%. However the latest figures from the City transport authority show this figure having reduced to 6%. Given the growth of the city in both population and spatial terms, this may reflect no real increase in the effective level of supply.

Whilst it is not yet possible to interrogate the registration database in any real detail, there is some indication that recent registrations have been the higher quality vehicles that are capable of being operated on the so-called ‘hot’ routes requiring a good hill-climbing capability. Even here, though, there is anecdotal evidence of a decline in values, as such vehicles had been reputedly trading at Birr 120,000 to 130,000 but now can command only some Birr 70,000 to 80,000 (say US$ 8,700). This lower value is typical of the region, and indicates that Addis Ababa can no longer justify a premium.

Taken together, all available indicators confirm the lack of profitability in the sector and hence its inability to attract appropriate investment.

Low skills base

A low skills base in almost all aspects of its operations typifies the urban passenger transport industry in Addis Ababa. As has been noted earlier, driving standards are low and there have been doubts about the testing and licensing programmes. However problems are not just confined to this segment of the business.

Business managers have a poor understanding of financial issues, and basically operate at the short-term cashflow level. Actions that might be regarded as investments offering a high rate of return, such as effective preventive maintenance, are generally ignored. No attempts are made to minimise life-cycle costs or to maximise returns on investment. Concepts such as amortisation are not understood.

At the operational level, decisions are made about vehicle despatch without consideration of any potential passengers that might be waiting along the route. Little differential pricing is used in order to attract passengers in the off-peak when there is a surplus of capacity, though it is reported that exploitation of the fares structure is less prevalent at these times.

Network planning is an ad hoc process based on the positioning of existing terminals. However relatively little network development takes place in this manner, despite the need displayed by the high rates of interchange that passengers are forced to make.

Finally, the sector has failed to produce any obvious entrepreneurs capable of differentiation or innovation in the services being provided. International experience indicates that there is a generally a market for premium services capable of attracting the marginal car user. These can be priced at a level ensuring a good return on capital employed.

The primary cause for these low skills is the lack of incentives for their improvement. Whilst the necessary regulatory framework is broadly in place, its enforcement is weak and patchy at best. Only when operators perceive real risks in their continuing practices will they endeavour to improve these. It has been suggested in the study on improving urban transport that this might be driven through the introduction of operator licensing, but this will only prove to be effective if the appropriate supporting institutional frameworks are put in place.

Violent or illegal behaviour in the operating industry, and administrative corruption

By the general standards of urban passenger transport operations in sub-Saharan Africa, those in Addis Ababa are models of good order typified by a lack of violence between operators. There is a degree of on-the-road competition between vehicles on common sections of route, but this rarely deteriorates beyond blocking practices at stops that are an inevitable consequence of the lack of appropriate infrastructure.

As the public authorities have abdicated virtually all responsibilities with regard to regulation of the industry, their probity in this regard is not an issue. Operating permits are made freely available to all potential new market entrants, and no reports have been alleged of these being withheld against a demand for personal payment.

What is of greater concern is the apparent level of administrative inefficiency, though this is now beginning to be addressed. Data on operating permits are not collected accurately, no one knows whether the fees cover the administrative costs involved in their issue, and again no one knows whether there is widespread evasion. However, given the very low charges, it seems unlikely that intending operators would risk not obtaining the appropriate permit.

Where problems do exist, these relate to road transport in general and not just to the urban passenger transport sector in particular. As has been noted earlier the Transport Authority lacks institutional capacity, and both it and the Police (Traffic Controllers) lack resources. Allegations of petty corruption had been made against both organisations, but clearly these are hard to substantiate.

General opinion suggests that the City administration, under new political and managerial leadership, has largely put its house in order. However feedback to the minibus owners associations from drivers suggests that there is still a degree of Police harassment. Where traffic offences are punishable by relatively small fines, but repeat offences incur licence suspensions, there is a strong incentive for drivers to offer an enticement to controllers whether or not these are demanded.

Lack of economic regulation, network planning, and supporting institutions

In effect, the 1992 Proclamation removed any form of economic regulation of the urban passenger transport system other than for the Road Transport Authority to propose fares that are then set by the Ministry. In all other respects the Proclamation relates to the ‘quality’ regulation of transport, setting the standards for vehicles and drivers and their registration / licensing. Any powers in respect of route planning relate only to inter-regional transport.

As has been noted earlier, the network of services provided through self-regulation is not an efficient match of transport supply to passenger demand. Further, the lack of real competition within the supply industry fails to exert control over tariffs or to act as an incentive to raise service quality. In effect, producer interests are given priority over those of the consumer.

International research in recent years has identified that neither regulation nor deregulation provide optimal frameworks for raising both the efficiency of production and the efficiency in consumption of urban passenger transport services. The study on improving urban transport argued strongly for the introduction of economic regulation for the sector, based on the principle of separating service planning from its actual delivery.

A route network is designed, the service levels over this are specified, and the right to operate these services is then offered under competitive tender. A combination of the relevant technical skills and the application of market forces reduce the resources consumed by the sector, and these can be redeployed either to reduce general prices, to increase the service offer, or to enable support to peripheral services through cross-subsidy.

However the local interpretation of these recommendations is that they should be restricted to the bus sector, and exclude minibuses. For the latter, it is planned to encourage more effective self-regulation through the strengthening of area-based owners associations and mandating that all drivers join one of these. Whilst this process has the potential to create a more orderly market, it will reduce the opportunity for competition in the sector. Further, unless the owner associations develop some hitherto hidden network planning capabilities, there are still likely to be major inefficiencies in this domain.

For the bus sector, there cannot be any competition for the provision of highly subsidised bus services unless the conditions of such subsidy are made available equitably to all potential bidders. Potentially this leaves the responsibility for all future capital investment resting with the public sector, whereas the mobilisation of private capital might be seen as the better option. However competition could still be brought into the operation of the vehicles through making them available on operating leases to potential route contractors. Competition for the supply of the vehicles could come through kilometre priced contracts for their provision and maintenance.

In the middle ground lies the potential for commercial bus services operated under a regime of limited competition. These could either be relatively standard services, as proposed in the study on improving transport, or genuinely premium services capable of attracting car users. In either case, the main problem will come in mobilising the necessary local capital given the lack of investor activity in the sector and the delays in creating capital markets.

For effective governance of the urban transport sector in Addis Ababa, an appropriate institutional framework needs to be put in place. This must cover both the planning and procurement of the transport services and their supporting infrastructure, and also the regulation of the actual service delivery. Whatever the preferred framework, new or revised primary legislation may well be required for the establishment of the relevant bodies.

Inappropriate ownership structures or company size

As has been noted earlier, the large majority of public transport vehicles in Addis Ababa are in single unit ownership with fleet sizes of four or more being extremely rare. Further, in many cases the vehicle owner is a semi-passive investor with no direct involvement in its day-to-day operation. His commitment to the industry is not strong, and he may withdraw from it when faced with major expenditure whether for technical or accident repairs. Not only does such an ownership pattern restrict the likely level of capital investment that may be made, but it also produces incentives for maximising short-term cashflows.

Investment issues are further affected by the availability of commercial finance for vehicle procurement. Such owners cannot produce a credible business plan, and generally have no or insufficient collateral to provide alternative security to lenders. Under these circumstances it is unsurprising that available finance terms are onerous, with high deposits, short terms, and interest rate risk premiums. As such, investment funds tend to be sourced informally through the extended family or credit unions (EKUB). Funding resources restrict both the size and the age of the vehicle that may then be purchased.

The emphasis on short-term cashflow for the owner results in the setting of daily hire rates that leave little margin for the driver after meeting his unavoidable costs. This, in turn, leads to incentives for speeding and aggressive driving in traffic and at stops in order to maximise passenger revenues. It also results in deferral of repair expenditure on the vehicle, where this is it all possible, and avoidance of preventive maintenance. This, in turn, explains the typical vehicle condition observed on the road.

In the particular case of Anbessa, this enterprise is still owned by the Federal government as a result of the failure to divest the component divisions of the Passenger Transport Corporation. However, apart from seven buses operated in the remote centre of Jimma, all of the activities of the enterprise are based in Addis Ababa. Should the enterprise remain in public ownership it would be logical to transfer responsibility to the City administration, but such a move would require a restructuring of its balance sheet.

At present this contains assets and liabilities relating both to the former Passenger Transport Corporation and to intra-government transfers such as duty payments. For the former there is inadequate documentation to enable these to be pursued / honoured, and for the latter their values would become crystallised at the point of transfer. Recommendations in this domain were made by the consultants employed by the Ethiopian Privatisation Agency when Anbessa was included in Tranche 3 of its programme, but have largely yet to be implemented.

Finally, it should also be noted that Government provides private commercial road transport for many of its officers on their commuting journeys. If an appropriate bus service were in place, then the provision of tickets for this would be a more efficient means of such income support. If contract services were genuinely required, then these would best be provided by private contractors.

Inadequate transport infrastructure

As has been noted in the previous, Chapter some roads in Addis Ababa are subject to severe congestion (at peak times in particular) resulting in low operating speeds for public transport vehicles as well as for general traffic. Whilst road capacity has been expanded on the main arterials and orbitals in recent years, there are still problems with intersection management and functionality of the design standards. In particular, very little provision has been made with respect to public transport stops so that vehicles usually have to stay in the carriageway while passengers board and alight.

The urban transport planning study will identify a number of low-cost measures that could be taken in the near term to improve traffic flows, as well as a range of longer-term schemes involving more significant investments. Taken together, these would raise general operating speeds and hence vehicle productivity in the public transport sector.

However the study will also recommend the introduction of public transport priority measures through the re-allocation of existing road space and the construction of additional capacity where needed. These actions are justified on the grounds of efficiency of utilisation of road space, with public transport vehicles carrying some three quarters of all people movements despite forming little more than one quarter of the traffic flow. Constraints on private car movements arising as a result can be seen as one action in a demand management strategy.

Improvements in vehicle productivity made possible by better transport infrastructure would also enable an improvement in fixed-cost recovery by operators and so encourage investment in more expensive but more efficient vehicle types. This would allow a polarisation of the transport supply, with both low-fare mass transit buses and premium-fare luxury midi-buses capable of attracting existing car users.

Lack of empowerment of transport users

In assessing the problems of the urban transport sector, it becomes increasingly clear that the voice of the consumer is not being heard. They have no representation on development bodies, and are powerless in the face of the operating industry effectively controlled by the marshals. As a result of the latter, they are forced to access services at points that may not be convenient and to use the next available vehicle irrespective of its condition. At times of lower demand, they may be forced to wait an unreasonable time until any service is offered at all.

The introduction of a degree of regulation into the urban passenger transport market would potentially redress the balance of power between the consumer and the producer. Not only could regulations be set in passenger interests, such as the maximum headways on routes, but passengers could also be empowered as a first-level enforcement agency.

In setting relevant standards, a fuller understanding of passenger priorities is required through an extension of the survey programme undertaken for the study on improving transport in both qualitative and spatial terms. Different issues will assume higher priorities in each locality, and will need to be addressed accordingly. Standards of service, and vehicle types, may need to vary to meet local constraints – particularly on affordability.

Were the Transport Authority to establish a formal complaints procedure, reports to this could have a direct impact on the license of the relevant operator or driver. Further, were any offences committed by the driver, imposition of penalties that would also affect the passengers (such as temporary impounding of the vehicle) would provide incentives for them to exercise a degree of control over his behaviour.

Reform programme

Nature of planned reforms

As a result of the study on improving transport in Addis Ababa, the City administration and its Transport Authority have started to plan a regulatory reform of the urban transport sector.

However, as discussed in Section 9.5, this reform embodies different approaches for the bus and minibus sectors. For the former, it is planned to introduce economic regulation for the competitive provision of services. For the latter, it is intended to enhance the self-regulatory powers of the owners’ associations.

It can be argued that these two approaches are contradictory, and therefore will not result in an effective and efficient integrated public transport system. In addition, the planned reform of the minibus sector (currently meeting nearly half of all public transport demand) will not provide for an enhanced network planning capability. In any case, current orthodoxy suggests that there should be a separation between the planning and production of transport services so as to provide incentives in this domain.

For the bus sector, there needs to be either a formal recognition that the exclusive franchise formerly held by Anbessa no longer applies or that it is confined to the provision of socially necessary services. The opening of any commercial bus services to competition will need to identify and foster potential new market entrants, and ensure genuine equality of opportunity. It is not clear that such conditions are yet in place.

Drivers of the proposed reforms

The main driver of the proposed reforms is the revived City administration under its new political leadership. The World Bank and PPIAF have given active support in funding the necessary sector studies.

Obstacles to reform

The two major stumbling blocks in introducing the proposed reform measures are the lack of an institutional framework within which to introduce these, and the potential opposition of the existing marshals and their controlling unions.

Tighter regulation of the minibus sector will require strengthening of the existing powers suggesting membership of an area-based association as a condition of an operating license, but the Transport Authority believes that it can issue a directive to that effect.

Of greater concern is that the proposed framework would make the marshals subservient to the new owners’ associations. At present, the balance of power between these two sectors is very much in favour of the former on account of their relative funds flow. Any unauthorised body controlling the marshals is unlikely to surrender its power willingly, though again the Transport Authority believe that has a legal framework to ensure this and can mobilise the Police to that effect.

Linkage to other development initiatives

At this point in time, there is no linkage between the proposed sector reforms and other development initiatives.

However a concept study has been carried out for a mass transit scheme on the main east-west corridor through the heart of the city. Whilst this was originally rail-based, its conversion to bus rapid transit might enable such a linkage.

Appendix 1

Anbessa Route Analysis – January 2005

Core Routes

|Passenger |Ticket |Buses |Distance |Route |Fare |Rev + |Actual |Bus |Trips |Pax |Pax |Sched |Trip |Head |Sched |Sched |Hours |Lost |Lost |Bus |Daily |Bus |Stop | |Route | |Revenue |Used |operated |length | |Sub |Trips |PVR |sched |per |per |bus |time |way |speed |bus |per bus |trips |trip |chg |kms |stops |per | | | |Birr | |km |km |Birr |Br/km | | |daily |trip |bus day |kms |min |min |km/h |hours |/day |/day |% | |/bus | |km | |2 |219,536 |109,768 |270 |23,543.1 |11.1 |0.50 |5.59 |2,121 |5 |84 |103.5 |1,463.6 |932.4 |50.0 |20.0 |13.3 |70.0 |14.0 |13.3 |15.8 |120 |157.0 |19 |1.71 | |3 |462,260 |231,130 |485 |47,530.8 |10.8 |0.50 |5.84 |4,401 |10 |200 |105.0 |1,540.9 |2,160.0 |40.0 |8.0 |16.2 |133.3 |13.3 |53.3 |26.7 |185 |158.4 |15 |1.39 | |4 |201,078 |150,809 |333 |38,082.2 |19.4 |0.75 |4.49 |1,963 |6 |71 |102.4 |1,117.1 |1,377.4 |69.0 |23.0 |16.9 |81.7 |13.6 |5.6 |7.8 |153 |211.6 |15 |0.77 | |5 |164,680 |82,340 |206 |21,450.3 |12.7 |0.50 |4.61 |1,689 |4 |68 |97.5 |1,372.3 |863.6 |50.0 |25.0 |15.2 |56.7 |14.2 |11.7 |17.2 |86 |178.8 |20 |1.57 | |6 |613,891 |214,862 |526 |47,856.6 |9.9 |0.35 |5.77 |4,834 |10 |200 |127.0 |2,046.3 |1,980.0 |40.0 |8.0 |14.9 |133.3 |13.3 |38.9 |19.4 |226 |159.5 |17 |1.72 | |7 |13,130 |4,596 |30 |2,307.4 |8.3 |0.35 |2.56 |278 |1 |14 |47.2 |437.7 |116.2 |30.0 |60.0 |16.6 |7.0 |7.0 |4.7 |33.8 |0 |76.9 |15 |1.81 | |8 |160,025 |56,009 |175 |15,632.2 |9.4 |0.35 |4.61 |1,663 |3 |69 |96.2 |1,778.1 |648.6 |37.0 |24.7 |15.2 |42.6 |14.2 |13.6 |19.7 |85 |173.7 |14 |1.49 | |9 |216,912 |54,228 |207 |16,891.8 |9.4 |0.25 |4.49 |1,797 |4 |80 |120.7 |1,807.6 |752.0 |43.0 |21.5 |13.1 |57.3 |14.3 |20.1 |25.1 |87 |140.8 |18 |1.91 | |10 |274,912 |137,456 |316 |32,321.5 |12.7 |0.50 |5.10 |2,545 |6 |102 |108.0 |1,527.3 |1,295.4 |50.0 |16.7 |15.2 |85.0 |14.2 |17.2 |16.8 |136 |179.6 |21 |1.65 | |11 |330,534 |82,634 |239 |18,326.4 |6.9 |0.25 |6.31 |2,656 |4 |104 |124.4 |2,754.5 |717.6 |30.0 |15.0 |13.8 |52.0 |13.0 |15.5 |14.9 |119 |152.7 |15 |2.17 | |12 |478,739 |167,559 |477 |40,837.5 |9.9 |0.35 |5.28 |4,125 |8 |160 |116.1 |1,994.7 |1,584.0 |40.0 |10.0 |14.9 |106.7 |13.3 |22.5 |14.1 |237 |170.2 |14 |1.41 | |13 |255,044 |63,761 |205 |16,785.4 |8.9 |0.25 |5.32 |1,886 |4 |80 |135.2 |2,125.4 |712.0 |42.0 |21.0 |12.7 |56.0 |14.0 |17.1 |21.4 |85 |139.9 |14 |1.57 | |14 |186,105 |93,053 |234 |23,652.9 |12.3 |0.50 |4.72 |1,923 |4 |70 |96.8 |1,550.9 |861.0 |50.0 |25.0 |14.8 |58.3 |14.6 |5.9 |8.4 |114 |197.1 |13 |1.06 | |15 |245,973 |61,493 |209 |16,417.4 |8.3 |0.25 |5.24 |1,978 |4 |88 |124.4 |2,049.8 |730.4 |37.0 |18.5 |13.5 |54.3 |13.6 |22.1 |25.1 |89 |136.8 |14 |1.69 | |16 |377,335 |94,334 |289 |21,093.0 |7.9 |0.25 |6.26 |2,670 |6 |120 |141.3 |2,096.3 |948.0 |40.0 |13.3 |11.9 |80.0 |13.3 |31.0 |25.8 |109 |117.2 |16 |2.03 | |17 |332,096 |116,234 |310 |25,743.9 |9.1 |0.35 |5.80 |2,829 |6 |108 |117.4 |1,845.0 |982.8 |45.0 |15.0 |12.1 |81.0 |13.5 |13.7 |12.7 |130 |143.0 |14 |1.54 | |18 |302,706 |105,947 |391 |23,754.2 |7.3 |0.35 |5.73 |3,254 |6 |109 |93.0 |1,681.7 |795.7 |45.0 |15.0 |9.7 |81.8 |13.6 |0.5 |0.5 |211 |132.0 |20 |2.74 | |19 |311,194 |155,597 |399 |41,224.2 |12.7 |0.50 |4.53 |3,246 |6 |130 |95.9 |1,728.9 |1,651.0 |40.0 |13.3 |19.1 |86.7 |14.4 |21.8 |16.8 |219 |229.0 |17 |1.34 | |20 |188,474 |47,119 |225 |12,707.1 |6.3 |0.25 |5.19 |2,017 |3 |78 |93.4 |2,094.2 |491.4 |31.0 |20.7 |12.2 |40.3 |13.4 |10.8 |13.8 |135 |141.2 |13 |2.06 | |21 |202,695 |50,674 |214 |10,862.5 |5.5 |0.25 |6.53 |1,975 |3 |83 |102.6 |2,252.2 |456.5 |30.0 |20.0 |11.0 |41.5 |13.8 |17.2 |20.7 |124 |120.7 |11 |2.00 | |23 |194,162 |97,081 |234 |23,212.8 |12.4 |0.50 |5.02 |1,872 |4 |64 |103.7 |1,618.0 |793.6 |50.0 |25.0 |14.9 |53.3 |13.3 |1.6 |2.5 |114 |193.4 |21 |1.69 | |24 |72,745 |54,559 |144 |13,699.8 |17.7 |0.75 |4.51 |774 |2 |27 |94.0 |1,212.4 |477.9 |60.0 |60.0 |17.7 |27.0 |13.5 |1.2 |4.4 |84 |228.3 |26 |1.47 | |25 |171,695 |128,771 |187 |27,645.0 |19.0 |0.75 |5.28 |1,455 |6 |78 |118.0 |953.9 |1,482.0 |60.0 |20.0 |19.0 |78.0 |13.0 |29.5 |37.8 |7 |153.6 |30 |1.58 | |26 |97,965 |122,456 |208 |32,206.5 |25.5 |1.25 |4.11 |1,263 |4 |50 |77.6 |816.4 |1,275.0 |61.0 |30.5 |25.1 |50.8 |12.7 |7.9 |15.8 |88 |268.4 |31 |1.22 | | |Passenger |Ticket |Buses |Distance |Route |Fare |Rev + |Actual |Bus |Trips |Pax |Pax |Sched |Trip |Head |Sched |Sched |Hours |Lost |Lost |Bus |Daily |Bus |Stop | |Route | |Revenue |Used |operated |length | |Sub |Trips |PVR |sched |per |per |bus |time |way |speed |bus |per bus |trips |trip |chg |kms |stops |per | | | |Birr | |km |km |Birr |Br/km | | |daily |trip |bus day |kms |min |min |km/h |hours |/day |/day |% | |/bus | |km | |27 |402,126 |100,532 |367 |29,699.0 |8.5 |0.25 |4.74 |3,494 |8 |174 |115.1 |1,675.5 |1,479.0 |38.0 |9.5 |13.4 |110.2 |13.8 |57.5 |33.1 |127 |123.7 |16 |1.88 | |28 |221,926 |110,963 |283 |24,497.7 |11.1 |0.50 |5.44 |2,207 |4 |79 |100.6 |1,849.4 |876.9 |42.0 |21.0 |15.9 |55.3 |13.8 |5.4 |6.9 |163 |204.1 |17 |1.53 | |29 |265,635 |132,818 |312 |29,006.8 |12.7 |0.50 |5.49 |2,284 |6 |93 |116.3 |1,475.8 |1,181.1 |55.0 |18.3 |13.9 |85.3 |14.2 |16.9 |18.1 |132 |161.1 |14 |1.10 | |30 |54,504 |68,130 |121 |14,860.8 |25.8 |1.25 |4.95 |576 |2 |20 |94.6 |908.4 |516.0 |60.0 |60.0 |25.8 |20.0 |10.0 |0.8 |4.0 |61 |247.7 |  |  | |31 |610,179 |152,545 |453 |35,053.8 |7.4 |0.25 |6.09 |4,737 |8 |213 |128.8 |2,542.4 |1,576.2 |30.0 |7.5 |14.8 |106.5 |13.3 |55.1 |25.9 |213 |146.1 |11 |1.49 | |32 |285,869 |100,054 |349 |28,948.6 |10.6 |0.35 |4.44 |2,731 |6 |120 |104.7 |1,588.2 |1,272.0 |41.0 |13.7 |15.5 |82.0 |13.7 |29.0 |24.1 |169 |160.8 |13 |1.23 | |33 |280,996 |140,498 |339 |30,723.0 |11.4 |0.50 |5.49 |2,695 |5 |93 |104.3 |1,873.3 |1,060.2 |43.0 |17.2 |15.9 |66.7 |13.3 |3.2 |3.4 |189 |204.8 |21 |1.84 | |34 |188,537 |65,988 |229 |18,512.2 |9.8 |0.35 |4.58 |1,889 |4 |72 |99.8 |1,571.1 |705.6 |45.0 |22.5 |13.1 |54.0 |13.5 |9.0 |12.5 |109 |154.3 |17 |1.73 | |35 |174,712 |43,678 |163 |10,840.6 |6.7 |0.25 |5.64 |1,618 |3 |70 |108.0 |1,941.2 |469.0 |35.0 |23.3 |11.5 |40.8 |13.6 |16.1 |23.0 |73 |120.5 |10 |1.49 | |36 |168,706 |84,353 |215 |20,287.8 |11.7 |0.50 |4.99 |1,734 |4 |65 |97.3 |1,405.9 |760.5 |51.0 |25.5 |13.8 |55.3 |13.8 |7.2 |11.1 |95 |169.1 |17 |1.45 | |37 |155,835 |77,918 |206 |19,596.0 |12.0 |0.50 |4.77 |1,633 |4 |64 |95.4 |1,298.6 |768.0 |52.0 |26.0 |13.8 |55.5 |13.9 |9.6 |14.9 |86 |163.3 |26 |2.17 | |38 |191,507 |95,754 |234 |22,022.0 |11.0 |0.50 |5.22 |2,002 |4 |78 |95.7 |1,595.9 |858.0 |45.0 |22.5 |14.7 |58.5 |14.6 |11.3 |14.4 |114 |183.5 |18 |1.64 | |39 |152,864 |76,432 |211 |17,673.6 |9.6 |0.50 |5.19 |1,841 |4 |68 |83.0 |1,273.9 |652.8 |40.0 |20.0 |14.4 |45.3 |11.3 |6.6 |9.8 |91 |147.3 |10 |1.04 | |40 |114,379 |85,784 |180 |18,061.1 |17.9 |0.75 |5.38 |1,009 |3 |36 |113.4 |1,270.9 |644.4 |65.0 |43.3 |16.5 |39.0 |13.0 |2.4 |6.6 |90 |200.7 |30 |1.68 | |41 |270,788 |67,697 |224 |18,275.0 |8.5 |0.25 |5.19 |2,150 |4 |80 |125.9 |2,256.6 |680.0 |40.0 |20.0 |12.8 |53.3 |13.3 |8.3 |10.4 |104 |152.3 |15 |1.76 | |42 |207,785 |72,725 |307 |22,099.0 |9.8 |0.35 |4.23 |2,255 |4 |95 |92.1 |1,731.5 |931.0 |35.0 |17.5 |16.8 |55.4 |13.9 |19.8 |20.9 |187 |184.2 |13 |1.33 | |43 |41,929 |62,894 |121 |14,314.8 |30.2 |1.50 |4.69 |474 |2 |16 |88.5 |698.8 |483.2 |90.0 |90.0 |20.1 |24.0 |12.0 |0.2 |1.3 |61 |238.6 |30 |0.99 | |44 |37,016 |55,524 |120 |14,500.8 |30.4 |1.50 |4.08 |477 |2 |16 |77.6 |616.9 |486.4 |88.0 |88.0 |20.7 |23.5 |11.7 |0.1 |0.6 |60 |241.7 |40 |1.32 | |45 |155,222 |77,611 |181 |15,746.6 |8.6 |0.50 |5.91 |1,831 |3 |72 |84.8 |1,724.7 |619.2 |35.0 |23.3 |14.7 |42.0 |14.0 |11.0 |15.2 |91 |175.0 |11 |1.28 | |46 |294,396 |147,198 |337 |32,132.8 |11.2 |0.50 |5.50 |2,869 |6 |99 |102.6 |1,635.5 |1,108.8 |50.0 |16.7 |13.4 |82.5 |13.8 |3.4 |3.4 |157 |178.5 |15 |1.34 | |47 |126,725 |44,354 |163 |9,015.3 |6.3 |0.35 |6.33 |1,431 |2 |48 |88.6 |2,112.1 |302.4 |35.0 |35.0 |10.8 |28.0 |14.0 |0.3 |0.6 |103 |150.3 |11 |1.75 | |48 |140,252 |70,126 |181 |18,246.6 |10.9 |0.50 |4.61 |1,674 |3 |61 |83.8 |1,558.4 |664.9 |41.0 |27.3 |16.0 |41.7 |13.9 |5.2 |8.5 |91 |202.7 |12 |1.10 | |49 |232,839 |58,210 |169 |14,382.0 |6.8 |0.25 |5.67 |2,115 |3 |108 |110.1 |2,587.1 |734.4 |21.0 |14.0 |19.4 |37.8 |12.6 |37.5 |34.7 |79 |159.8 |11 |1.62 | |50 |155,902 |77,951 |199 |18,658.2 |12.1 |0.50 |5.01 |1,542 |3 |54 |101.1 |1,732.2 |653.4 |45.0 |30.0 |16.1 |40.5 |13.5 |2.6 |4.8 |109 |207.3 |17 |1.40 | |51 |173,986 |43,497 |159 |12,595.5 |8.1 |0.25 |4.83 |1,555 |3 |67 |111.9 |1,933.2 |542.7 |37.0 |24.7 |13.1 |41.3 |13.8 |15.2 |22.6 |69 |140.0 |15 |1.85 | |52 |300,454 |150,227 |420 |37,181.7 |14.1 |0.50 |4.85 |2,637 |6 |82 |113.9 |1,669.2 |1,156.2 |59.0 |19.7 |14.3 |80.6 |13.4 |-5.9 |-7.2 |240 |206.6 |23 |1.63 | |53 |147,601 |73,801 |171 |18,296.5 |11.5 |0.50 |4.84 |1,591 |3 |59 |92.8 |1,640.0 |678.5 |42.0 |28.0 |16.4 |41.3 |13.8 |6.0 |10.1 |81 |203.3 |18 |1.57 | |54 |189,789 |94,895 |226 |17,793.5 |9.5 |0.50 |6.40 |1,873 |4 |68 |101.3 |1,581.6 |646.0 |48.0 |24.0 |11.9 |54.4 |13.6 |5.6 |8.2 |106 |148.3 |19 |2.00 | |55 |98,388 |49,194 |119 |11,485.5 |9.5 |0.50 |5.14 |1,209 |2 |48 |81.4 |1,639.8 |456.0 |35.0 |35.0 |16.3 |28.0 |14.0 |7.7 |16.0 |59 |191.4 |14 |1.47 | |56 |157,409 |78,705 |182 |19,397.2 |14.2 |0.50 |4.87 |1,366 |3 |46 |115.2 |1,749.0 |653.2 |54.0 |36.0 |15.8 |41.4 |13.8 |0.5 |1.0 |92 |215.5 |25 |1.76 | | |Passenger |Ticket |Buses |Distance |Route |Fare |Rev + |Actual |Bus |Trips |Pax |Pax |Sched |Trip |Head |Sched |Sched |Hours |Lost |Lost |Bus |Daily |Bus |Stop | |Route | |Revenue |Used |operated |length | |Sub |Trips |PVR |sched |per |per |bus |time |way |speed |bus |per bus |trips |trip |chg |kms |stops |per | | | |Birr | |km |km |Birr |Br/km | | |daily |trip |bus day |kms |min |min |km/h |hours |/day |/day |% | |/bus | |km | |57 |205,074 |102,537 |242 |24,710.4 |14.4 |0.50 |4.98 |1,716 |4 |61 |119.5 |1,709.0 |878.4 |54.0 |27.0 |16.0 |54.9 |13.7 |3.8 |6.2 |122 |205.9 |22 |1.53 | |58 |32,712 |21,263 |60 |5,940.0 |12.0 |0.65 |4.13 |495 |2 |16 |66.1 |545.2 |192.0 |45.0 |45.0 |16.0 |12.0 |6.0 |-0.5 |-3.1 |0 |99.0 |16 |1.33 | |59 |198,794 |99,397 |227 |23,276.0 |11.5 |0.50 |5.12 |2,024 |4 |88 |98.2 |1,656.6 |1,012.0 |37.5 |18.8 |18.4 |55.0 |13.8 |20.5 |23.3 |107 |194.0 |17 |1.48 | |60 |109,644 |246,699 |210 |56,404.0 |47.2 |2.25 |4.57 |1,195 |5 |50 |91.8 |731.0 |2,360.0 |80.0 |32.0 |35.4 |66.7 |13.3 |10.2 |20.3 |60 |376.0 |46 |0.97 | |61 |174,448 |87,224 |209 |21,735.0 |13.8 |0.50 |4.82 |1,575 |3 |55 |110.8 |1,938.3 |759.0 |45.0 |30.0 |18.4 |41.3 |13.8 |2.5 |4.5 |119 |241.5 |22 |1.59 | |62 |53,992 |67,490 |90 |18,988.0 |23.5 |1.25 |3.84 |808 |2 |26 |66.8 |899.9 |611.0 |60.0 |60.0 |23.5 |26.0 |13.0 |-0.9 |-3.6 |30 |316.5 |28 |1.19 | |63 |122,913 |43,020 |155 |12,749.1 |9.1 |0.35 |4.34 |1,401 |3 |70 |87.7 |1,365.7 |637.0 |35.0 |23.3 |15.6 |40.8 |13.6 |23.3 |33.3 |65 |141.7 |14 |1.54 | |64 |210,994 |105,497 |282 |22,021.0 |9.5 |0.50 |5.75 |2,318 |5 |87 |91.0 |1,406.6 |826.5 |50.0 |20.0 |11.4 |72.5 |14.5 |9.7 |11.2 |132 |146.8 |13 |1.37 | |65 |89,138 |57,940 |124 |12,155.0 |11.0 |0.65 |5.50 |1,105 |2 |44 |80.7 |1,485.6 |484.0 |40.0 |40.0 |16.5 |29.3 |14.7 |7.2 |16.3 |64 |202.6 |17 |1.55 | |66 |206,548 |103,274 |283 |21,294.0 |10.5 |0.50 |5.82 |2,028 |4 |68 |101.8 |1,721.2 |714.0 |47.0 |23.5 |13.4 |53.3 |13.3 |0.4 |0.6 |163 |177.5 |14 |1.33 | |67 |191,527 |95,764 |219 |22,266.6 |10.2 |0.50 |5.16 |2,183 |4 |88 |87.7 |1,596.1 |897.6 |37.0 |18.5 |16.5 |54.3 |13.6 |15.2 |17.3 |99 |185.6 |16 |1.57 | |68 |271,422 |67,856 |233 |19,383.6 |8.7 |0.25 |4.90 |2,228 |4 |82 |121.8 |2,261.9 |713.4 |40.0 |20.0 |13.1 |54.7 |13.7 |7.7 |9.4 |113 |161.5 |11 |1.26 | |69 |14,245 |7,123 |32 |2,679.0 |11.4 |0.50 |3.19 |235 |2 |20 |60.6 |237.4 |228.0 |45.0 |45.0 |15.2 |15.0 |7.5 |12.2 |60.8 |-28 |44.7 |19 |1.67 | |70 |187,678 |93,839 |237 |24,134.0 |11.0 |0.50 |4.67 |2,194 |4 |80 |85.5 |1,564.0 |880.0 |40.0 |20.0 |16.5 |53.3 |13.3 |6.9 |8.6 |117 |201.1 |16 |1.45 | |71 |124,483 |62,242 |178 |15,507.8 |10.6 |0.50 |4.82 |1,463 |3 |50 |85.1 |1,383.1 |530.0 |50.0 |33.3 |12.7 |41.7 |13.9 |1.2 |2.5 |88 |172.3 |19 |1.79 | |72 |109,426 |38,299 |141 |10,828.8 |9.6 |0.35 |4.55 |1,128 |3 |50 |97.0 |1,215.8 |480.0 |49.0 |32.7 |11.8 |40.8 |13.6 |12.4 |24.8 |51 |120.3 |18 |1.88 | |73 |209,064 |104,532 |225 |21,516.7 |10.3 |0.50 |5.83 |2,089 |4 |80 |100.1 |1,742.2 |824.0 |42.0 |21.0 |14.7 |56.0 |14.0 |10.4 |13.0 |105 |179.3 |18 |1.75 | |74 |176,101 |88,051 |207 |18,979.1 |13.3 |0.50 |5.57 |1,427 |4 |56 |123.4 |1,467.5 |744.8 |55.0 |27.5 |14.5 |51.3 |12.8 |8.4 |15.1 |87 |158.2 |19 |1.43 | |75 |166,262 |83,131 |237 |22,245.6 |10.4 |0.50 |4.48 |2,139 |4 |81 |77.7 |1,385.5 |842.4 |40.0 |20.0 |15.6 |54.0 |13.5 |9.7 |12.0 |117 |185.4 |15 |1.44 | |76 |149,061 |74,531 |173 |16,673.8 |12.1 |0.50 |5.36 |1,378 |3 |49 |108.2 |1,656.2 |592.9 |50.0 |33.3 |14.5 |40.8 |13.6 |3.1 |6.3 |83 |185.3 |21 |1.74 | |77 |219,583 |109,792 |239 |26,117.0 |13.0 |0.50 |5.04 |2,009 |4 |72 |109.3 |1,829.9 |936.0 |45.0 |22.5 |17.3 |54.0 |13.5 |5.0 |7.0 |119 |217.6 |21 |1.62 | |78 |25,571 |12,786 |41 |3,930.8 |12.4 |0.50 |3.90 |317 |2 |20 |80.7 |426.2 |248.0 |50.0 |50.0 |14.9 |16.7 |8.3 |9.4 |47.2 |-19 |65.5 |20 |1.61 | |79 |182,475 |91,238 |233 |22,446.6 |11.4 |0.50 |4.88 |1,969 |4 |72 |92.7 |1,520.6 |820.8 |45.0 |22.5 |15.2 |54.0 |13.5 |6.4 |8.8 |113 |187.1 |19 |1.67 | |80 |258,311 |129,156 |293 |32,364.0 |12.4 |0.50 |4.79 |2,610 |5 |88 |99.0 |1,722.1 |1,091.2 |50.0 |20.0 |14.9 |73.3 |14.7 |1.0 |1.1 |143 |215.8 |19 |1.53 | |81 |100,685 |50,343 |125 |12,765.2 |9.7 |0.50 |4.73 |1,316 |2 |48 |76.5 |1,678.1 |465.6 |35.0 |35.0 |16.6 |28.0 |14.0 |4.1 |8.6 |65 |212.8 |17 |1.75 | |82 |163,530 |81,765 |180 |15,808.7 |11.3 |0.50 |6.21 |1,399 |3 |48 |116.9 |1,817.0 |542.4 |53.0 |35.3 |12.8 |42.4 |14.1 |1.4 |2.8 |90 |175.7 |19 |1.68 | |83 |106,299 |53,150 |122 |14,752.2 |13.8 |0.50 |4.32 |1,069 |2 |36 |99.4 |1,771.7 |496.8 |47.0 |47.0 |17.6 |28.2 |14.1 |0.4 |1.0 |62 |245.9 |22 |1.59 | |84 |194,121 |48,530 |172 |10,406.4 |6.4 |0.25 |6.53 |1,626 |3 |68 |119.4 |2,156.9 |435.2 |36.0 |24.0 |10.7 |40.8 |13.6 |13.8 |20.3 |82 |115.6 |13 |2.03 | |85 |37,484 |84,339 |91 |17,820.0 |45.0 |2.25 |4.94 |396 |2 |16 |94.7 |624.7 |720.0 |99.0 |99.0 |27.3 |26.4 |13.2 |2.8 |17.5 |31 |297.0 |34 |0.76 | |87 |109,425 |54,713 |122 |13,608.0 |10.5 |0.50 |4.82 |1,296 |2 |37 |84.4 |1,823.8 |388.5 |35.0 |35.0 |18.0 |21.6 |10.8 |-6.2 |-16.8 |62 |226.8 |31 |2.95 | | |Passenger |Ticket |Buses |Distance |Route |Fare |Rev + |Actual |Bus |Trips |Pax |Pax |Sched |Trip |Head |Sched |Sched |Hours |Lost |Lost |Bus |Daily |Bus |Stop | |Route | |Revenue |Used |operated |length | |Sub |Trips |PVR |sched |per |per |bus |time |way |speed |bus |per bus |trips |trip |chg |kms |stops |per | | | |Birr | |km |km |Birr |Br/km | | |daily |trip |bus day |kms |min |min |km/h |hours |/day |/day |% | |/bus | |km | |88 |4,657 |13,971 |30 |2,640.0 |44.0 |3.00 |5.47 |60 |1 |2 |77.6 |155.2 |88.0 |90.0 |180.0 |29.3 |3.0 |3.0 |0.0 |0.0 |0 |88.0 |26 |0.59 | |89 |3,794 |11,382 |30 |2,640.0 |44.0 |3.00 |4.46 |60 |1 |2 |63.2 |126.5 |88.0 |90.0 |180.0 |29.3 |3.0 |3.0 |0.0 |0.0 |0 |88.0 |44 |1.00 | |90 |145,880 |72,940 |178 |17,440.0 |10.0 |0.50 |5.02 |1,744 |3 |60 |83.6 |1,620.9 |600.0 |40.0 |26.7 |15.0 |40.0 |13.3 |1.9 |3.1 |88 |193.8 |21 |2.10 | |92 |118,875 |59,438 |120 |10,704.0 |9.6 |0.50 |6.66 |1,115 |2 |40 |106.6 |1,981.3 |384.0 |43.0 |43.0 |13.4 |28.7 |14.3 |2.8 |7.1 |60 |178.4 |19 |1.98 | |93 |122,253 |61,127 |132 |13,872.0 |10.2 |0.50 |5.29 |1,360 |2 |55 |89.9 |2,037.6 |561.0 |30.0 |30.0 |20.4 |27.5 |13.8 |9.7 |17.6 |72 |231.2 |12 |1.18 | | | | | | | | | | | | | | | | | | | | | | | | | | | |Core |16,868,611 |7,626,888 |19,326 |1,818,491 |13.2 |0.60 |5.12 |162,188 |344 |6,366 |104.0 |1,634.6 |70,773.0 |47.2 |31.8 |16.8 |51.8 |13.4 |959.7 |15.1 |9,006 |177.3 |18.8 |1.57 | | | | | | | | | | | | | | | | | | | | | | | | | | | |Peak |153,686 |111,840 |383 |31,514 |10.9 |0.53 |4.04 |1,765 | | |87.1 | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |Total |17,022,297 |7,738,729 |19,709 |1,850,005 |12.7  |0.58  |5.10 |163,953 | | |103.8 | | | | | | | | | | | | | | |

Peak Supplements

|Passenger |Ticket |Buses |Distance |Route |Fare |Rev + |Actual |Bus |Trips |Pax | | | | | | | | | | | | | | |Route | |Revenue |Used |operated |length | |Sub |Trips |PVR |sched |per | | | | | | | | | | | | | | | | |Birr | |km |km |Birr |Br/km | | |daily |trip | | | | | | | | | | | | | | |6P |43 |15 |1 |15.8 |7.9 |0.35 |1.22 |2 | | |21.5 | | | | | | | | | | | | | | |11P |2,678 |670 |2 |80.0 |4.0 |0.25 |11.72 |20 | | |133.9 | | | | | | | | | | | | | | |12P |1,690 |592 |13 |68.4 |3.8 |0.35 |11.12 |18 | | |93.9 | | | | | | | | | | | | | | |17P |1,687 |590 |2 |129.6 |8.1 |0.35 |5.86 |16 | | |105.4 | | | | | | | | | | | | | | |21P |4,449 |1,112 |13 |188.0 |4.7 |0.25 |8.28 |40 | | |111.2 | | | | | | | | | | | | | | |25P |66,721 |66,721 |137 |16,416.0 |24.0 |1.00 |4.47 |684 | | |97.5 | | | | | | | | | | | | | | |32P |5,400 |1,890 |15 |364.0 |7.0 |0.35 |6.68 |52 | | |103.8 | | | | | | | | | | | | | | |33P |695 |348 |4 |70.0 |10.0 |0.50 |5.96 |7 | | |99.3 | | | | | | | | | | | | | | |40P |495 |371 |1 |70.8 |17.7 |0.75 |5.94 |4 | | |123.8 | | | | | | | | | | | | | | |42P |2,833 |992 |23 |280.0 |8.0 |0.35 |4.55 |35 | | |80.9 | | | | | | | | | | | | | | |46P |288 |144 |3 |15.0 |5.0 |0.50 |11.52 |3 | | |96.0 | | | | | | | | | | | | | | |47P |2,210 |774 |11 |132.5 |5.3 |0.35 |7.51 |25 | | |88.4 | | | | | | | | | | | | | | |49P |64 |16 |1 |4.0 |4.0 |0.25 |5.60 |1 | | |64.0 | | | | | | | | | | | | | | |50P |127 |64 |1 |12.1 |12.1 |0.50 |6.30 |1 | | |127.0 | | | | | | | | | | | | | | |52P |9,154 |4,577 |13 |834.3 |10.3 |0.50 |6.58 |81 | | |113.0 | | | | | | | | | | | | | | |57P |465 |233 |1 |92.4 |13.2 |0.50 |3.02 |7 | | |66.4 | | | | | | | | | | | | | | |59P |2,805 |1,403 |4 |280.8 |7.8 |0.50 |5.99 |36 | | |77.9 | | | | | | | | | | | | | | |61P |14,156 |7,078 |20 |1,600.0 |12.8 |0.50 |5.31 |125 | | |113.2 | | | | | | | | | | | | | | |66P |5,823 |2,912 |9 |682.5 |10.5 |0.50 |5.12 |65 | | |89.6 | | | | | | | | | | | | | | |69P |10,544 |5,272 |21 |1,630.2 |11.4 |0.50 |3.88 |143 | | |73.7 | | | | | | | | | | | | | | |73P |788 |394 |2 |76.5 |8.5 |0.50 |6.18 |9 | | |87.6 | | | | | | | | | | | | | | |77P |1,327 |664 |2 |143.0 |13.0 |0.50 |5.57 |11 | | |120.6 | | | | | | | | | | | | | | |78P |12,519 |6,260 |20 |1,798.0 |12.4 |0.50 |4.18 |145 | | |86.3 | | | | | | | | | | | | | | |84P |675 |169 |1 |44.8 |6.4 |0.25 |5.27 |7 | | |96.4 | | | | | | | | | | | | | | |93P |491 |246 |3 |30.6 |10.2 |0.50 |9.63 |3 | | |163.7 | | | | | | | | | | | | | | |26U |2,874 |4,311 |31 |3,210.0 |30.0 |1.50 |1.43 |107 | | |26.9 | | | | | | | | | | | | | | |62U |2,685 |4,028 |29 |3,245.0 |27.5 |1.50 |1.32 |118 | | |22.8 | | | | | | | | | | | | | | |

Appendix 2

Anbessa Financial Statements – Management Accounts

Ethiopian Calendar |1987 |1988 |1989 |1990 |1991 |1992 |1993 |1994 |1995 |1996 | |Gregorian Calendar |1994/95 |1995/96 |1996/97 |1997/98 |1998/99 |1999/2000 |2000/01 |2001/02 |2002/03 |2003/04 | | | | | | | | | | | | | |REVENUE |25,642 |18,539 |56,618 |93,588 |113,954 |120,242 |123,568 |116,995 |120,599 |130,884 | |Normal service |22,180 |15,237 |27,737 |46,078 |60,462 |65,678 |74,522 |73,033 |92,107 |100,704 | |Contract |3,193 |3,129 |2,661 |1,884 |1,896 |2,585 |4,193 |4,541 |3,245 |2,609 | |Subsidy | | |25,894 |45,153 |50,089 |49,134 |42,972 |36,645 |23,104 |25,876 | |Other |269 |173 |326 |473 |1,507 |2,845 |1,881 |2,776 |2,143 |1,695 | | | | | | | | | | | | | |EXPENDITURE |28,423 |34,736 |60,067 |89,755 |108,106 |117,144 |121,162 |124,437 |141,461 |169,402 | |Wages and benefits |10,153 |10,468 |12,611 |15,823 |18,696 |20,598 |24,259 |25,960 |29,876 |31,409 | |Fuel and lubricants |7,431 |6,283 |10,763 |15,739 |18,798 |22,950 |29,653 |30,655 |35,406 |41,518 | |Maintenance |3,260 |6,509 |4,209 |8,487 |15,788 |16,799 |19,875 |20,379 |26,028 |27,436 | |Spare parts |2,507 |5,810 |3,123 |5,737 |11,655 |10,706 |14,179 |13,596 |17,933 |20,950 | |Tyres & tubes |753 |699 |1,086 |2,750 |4,133 |6,093 |5,696 |6,783 |8,095 |6,486 | |Finance expenses |6,521 |9,449 |29,751 |45,915 |48,452 |47,292 |39,108 |39,621 |43,100 |59,059 | |Depreciation |6,519 |8,829 |26,525 |39,070 |42,208 |41,377 |39,108 |39,621 |43,100 |59,059 | |Interest |2 |620 |3,226 |6,845 |6,244 |5,915 |0 |0 |0 |0 | |Other expenses |1,058 |2,027 |2,733 |3,791 |6,372 |9,505 |8,267 |7,822 |7,051 |9,980 | |Insurance |0 |419 |604 |1,296 |2,235 |1,703 |1,779 |1,908 |2,577 |3,548 | |Printing & stationery |760 |765 |1,440 |2,135 |2,456 |3,344 |3,056 |3,037 |2,435 |2,953 | |Administration |209 |209 |211 |337 |534 |923 |480 |578 |807 |686 | |Miscellaneous |89 |634 |478 |23 |1,147 |3,535 |2,952 |2,299 |1,232 |2,793 | | | | | | | | | | | | | |NET INCOME |-2,781 |-16,197 |-3,449 |3,833 |5,848 |3,098 |2,406 |-7,442 |-20,862 |-38,518 | | | | | | | | | | | | | |Internal cash generation |3,738 |-7,368 |-2,818 |-2,250 |-2,033 |-4,659 |-1,458 |-4,466 |-866 |-5,335 | |Cashflow including subsidy |3,738 |-7,368 |23,076 |42,903 |48,056 |44,475 |41,514 |32,179 |22,238 |20,541 | |