In a developing economy, the public sector allocation will vary according to type
of service and infrastructure needs ([3]). Public services which register relatively
higher average access rates and are less dependent on a pre-existing infrastructure
grid can be envisaged to follow a more balanced diffusion process and be largely
pro-poor even in the absence of outright pro-poor policies, except if public authorities
pursue anti-egalitarian strategies. This pattern can be illustrated by primary
schools, which do not require substantial infrastructure investment in remote districts.
When they require instead substantial infrastructure and severely fall short of
universal coverage, public utilities expand more easily in and around districts which
are already partly endowed with the needed infrastructure (unless equalisation of
service provision is striven for as an urgent priority). In the medium term, service
improvements may turn out to remain skewed in favour of the non-poor, even though
social policies are not openly biased against the poor: maximising local service access
is less costly when delivery expansion is close to pre-existing infrastructure, where
residents exercise pressure for adequate service coverage and quality. Healthcare is
likely to follow this second pattern, with sparsely populated, poor districts often
benefiting marginally from public investment. For some public services, costly coping
mechanisms (e.g., water from trucks supplementing piped water supply; Walker
et al. [37]) are often the only option for many residents in areas with poor infrastructural
coverage. This typically concerns basic services such as telephone, water
and electricity, and is regarded by some authors as evidence against the rationale for
low user charges. In this view, rationing of excess demand might restrict subsidised
services to richer households, and can be redressed by adopting high user charges
coupled with discriminatory pricing (Thobani [33]; among counterarguments in recent
years, see Fredriksen [19]).