The A1 Four Lane Project: Will this be another PPP failure or is there a way forward?

DOUGLAS RASBASH* examines the viability and finds little that is attractive, hence docking at an almost completely different proposition


In 2001, there was almost no dual carriageway on any section of the A1 Highway. Now, less than 25 years later, there are plans to dual all 650km of it from the South African border at Ramatlabama to the border with Zimbabwe at Ramokgwebana.


The explosion in road traffic continues unabated with 640,000 registered vehicles in 2023 compared to 140,000 in 2001. Whether this is good economically is another debate because if traffic grows faster than the economy, it is not sustainable and public transport becomes less viable.


Botswana now has an enviable 20,000 km of paved roads, which is a remarkable achievement. But even with the highest number of vehicles per capita in Africa, that only gives an ownership per km density of 320 vehicles per paved km.


Of course, there are sections near the main cities where there are high traffic volumes, but that occurs only for 90 minutes in the morning and early evening. So the pressing question is whether there is sufficient traffic to support PPP.

Let us appreciate that all road investment so far has been funded directly by the government, which is rather relaxed on the issue of economics. But for private investors, economics is crucial. For toll roads to be feasible, the conventional wisdom suggests that 10,000 vehicles per day are needed.

Annual Average Daily Traffic

This would represent the Annual Average Daily Traffic or AADT through its entire length. Given the high density of urban traffic, the average trip length might only be 30km. PPP partners will be looking for 250km trip lengths to make it all worthwhile.


Furthermore, all of Botswana’s primary roads are mixed use, with lots of minor intersections, cattle and vendors at bus laybys and even veterinary checks. They cannot offer the same exclusive level of service as the SANRA network in South Africa.


Another factor is that road users have a choice between the toll road and other roads, this is universally taken as best practice. This is because all road users pay taxes and levies and are entitled to use roads without additional charges. With the A1 four-lane, there will be no such choice. So the issue is not only whether politically the project will be popular but that it is legally challengeable.


The last point is environmental, the government must know by now that road traffic emits considerable volumes of GHGs – greenhouse gases – which contributes to global warming. Building the four-lane does not help the feasibility of the environmentally more benign rail transport.


Let us look at the economics. Firstly, P20 billion or $1.4 billion significantly underestimates the cost for 630km of four lanes. It works out to be $2.2m per km. To illustrate this, the 23 km A1 Game City – Boatle four-lane cost P1.1 billion in 2017. Allowing for inflation, it will cost P1.5 billion today. That is P65 million per km or $4.7 million.


Remember there are no toll plazas. A more realistic estimate for the A1 project would be 630 km x $5 million per km or $3.15 billion, which is P43.5 billion. Now let us peep into the financials. To repay pula P43.5 billion over 20 years with a typical target ROI return on investment of 15%, an annual income of approximately P2.25 billion will be needed.

3 million vehicles km per day

Let us add 10 % for maintenance, administration and so forth, the annual income for the PPP concessionaire will be aiming for at least P 2.5 billion per year. Moving onto estimation of the toll, let us actually assume the target AADT of 10,000 travelling an average distance of 300 km – which is nowhere the case. That is 3 million vehicles km per day of about one billion veh km per year.

This means at a simple level, the toll will need setting at P2.5 per km. A more realistic assumption might be if the AADT is optimistically 5,000 travelling an average of 150 km, then the toll will need to be a whopping P10 per km. This means that a car travelling from Gaborone to Francistown would be charged P4,500.


In addition, the cost of fuel for the 450km trip will be about P500, making the total out-of-pocket expenditure a whopping P5,000, which is totally out of the question. Note that a single air trip costs less at P1230; the train ride used to be P90 and a bus is currently P155. It is blindingly obvious that the market will not accept even a break-even toll needed to travel by road from Gaborone to Francistown.

A further point is this: the government also would like heavy trucking and even some passenger trips to shift modes from road to rail, and has even built a rail line on the Kazungula bridge that will one day be linked with the railhead in Francistown. But heavy trucking would also be expected to provide a significant income to the proposed A1 four-lane road concession.

The government cannot expect both – there is confusion here. All potential concessionaires will understand this – the financial feasibility and the concomitant risks when the client is unclear and ambiguous about its intentions and is unconcerned about the projected economic feasibility.


The government should not be surprised to find that private sector interest in the project is not as enthusiastic as its own. But all is not lost because the expense of toll collection can be avoided as would be the economic cost of stopping traffic at toll plazas. If the cost of toll collection can be minimised, then the costs of the project capital and recurrent costs will reduce and its feasibility will be more positive.


However, it will take a very bold and visionary government to take the steps needed. The way around the problem of commercialisation of road space is to pay-as-you-drive.


This will require road users to have an on-board-unit or OBU (shown) that meters the distance the vehicle travels and charges accordingly. Such systems are commonly known as satellite tolling.

Geographical Navigation Satellite Systems or GNSS would know the precise location of the vehicle and its type, the road classification and quality, the time of day and much more. It can also label roads as being private of public, urban or rural and attribute the income from pay-as-you-drive accordingly to the riparian entity.

Kazungula Bridge

Even the toll plazas on the Kazungula Bridge can be done away with. It is an unavoidable fact that sooner or later governments will have to replace the old fuel-linked Road Fund Levy because of electric vehicles.


Some South African truckers are already showing interest in electric vehicles by Tesla Volvo and Mercedes that will be coming on the market from 2025. Such forward thinking companies will be able to avoid all fuel-based taxation.

If the government positions itself early on in the process, it can become a huge source of income, not only nationally but regionally too as other SADCC countries will be faced with similar circumstances. But will it do so?