Two Candidate Rural Roads and a Fixed Budget
A development fund can finance one rural road upgrade. Option A is 38 km of earth track in a township in the Chin Hills, serving about 9,000 people in 22 villages. It is impassable for roughly three months a year and passable only by four-wheel drive for a further two months. Farm incomes are low - terraced maize and beans, with most produce carried on foot to a market five hours away - but poverty incidence is high and there is no all-season access to a township hospital. Option B is 22 km of gravel road in an Ayeyarwady Region township serving about 34,000 people across flat paddy land. It floods for about three weeks in a normal monsoon year, and the villages already have a river ferry alternative that reaches the same market in 90 minutes. Option B has a benefit–cost ratio of 1.9 and Option A 1.2, because traffic counts on A are low.
1. Compare the population served per kilometre for the two options and explain why this indicator does not settle the choice.
Read the worked answer
Population per kilometre of road:
- Option A: 9,000 ÷ 38 km = 237 people per km.
- Option B: 34,000 ÷ 22 km = 1,545 people per km.
Option B serves about 6.5 times as many people per kilometre, and that is a legitimate efficiency argument. But the indicator measures the size of the served population, not the size of the change in their circumstances.
For Option B, villagers already reach the market by ferry in 90 minutes, so the road improves an existing connection: faster, cheaper and usable in some weather, but the marginal gain is modest and much of the traffic it carries already exists. For Option A, the villages have no all-season alternative at all, so the road converts a five-hour head-load journey - and three months of virtual isolation - into routine access. The gain per person is far larger even though fewer people receive it.
Population per kilometre also ignores who the users are. On A, the beneficiaries include patients who cannot reach a hospital in the monsoon and farmers who must sell at the farm gate because produce cannot be moved. Those losses never appear in a traffic count.
2. Which option would you fund, and what would have to be true for your answer to reverse?
Read the worked answer
I would fund Option A, but conditionally and with explicit reasoning. Both options return more than their cost (1.9 and 1.2), so this is not a choice between a good and a bad project but between two good ones. The case for A rests on there being no substitute: it buys access for a population that currently has none, and the benefits are concentrated on health access, price of marketed produce, and school attendance during the impassable months - outcomes the benefit–cost ratio handles poorly because it is built from observed traffic.
The answer should reverse to Option B under any of three conditions. First, if A's construction and maintenance risk is high: landslide repair on a hill road can consume 10–15% of capital cost annually, and if maintenance funding is not guaranteed, the road degrades and the benefit evaporates. Second, if a cheaper package - an upgraded track, footbridges, a scheduled ambulance and a health outreach team - delivers most of the access benefit for around a third of the cost. Third, if the fund's mandate is strictly to maximise measured economic return, in which case 1.9 must be preferred to 1.2.
3. Why might a national planner weight accessibility in the Chin Hills more heavily than the benefit–cost ratio suggests?
Read the worked answer
Because the benefit–cost ratio only counts things that can be priced from observed trips, and it treats existing traffic as the measure of value. In the Chin Hills the losses caused by isolation are largely invisible to that method: a mother who does not reach hospital in time, a child who misses school for three months, a maize crop sold at a depressed farm-gate price because it cannot be carried out before it spoils, a teacher who refuses a posting to a village with no dry-season road. None of these generates a traffic count.
There are also policy weights that a national planner legitimately applies and a project appraiser does not. Equity: the fund exists in part to reduce the gap between the well-connected delta and the hill townships, and a ratio that already reflects past investment will keep favouring already-served areas. Regional balance and security: an all-season link to a remote township supports administration, presence and service delivery in a border region. Irreversibility: a road, once built and maintained, allows future options - clinics, schools, markets, tourism - that cannot be forecast now, whereas the ferry villages already have optionality.
A planner should say this explicitly, as a political and distributional judgement layered on the economic one, rather than pretending the ratio has settled it.