. Estimates of the actual inci-
dence of government health spending are available
for only the late 1970s and beyond. Although these
estimates are not strictly comparable, because they
were computed by different authors using differ-
ent methods, they suggest that the targeting of
government health spending was quite pro-poor
in the late 1970s, then became less so during the
1980s to 1990s (table 7). In 2003/04 the poorest
Table 7 Sri Lanka: Incidence of public health expenditures, 1979-2004
Year
Share of government health expenditure
received by poorest household quintile (%)
Share of government health expenditure
received by poorest household quintile (%)
1979
30
9
1992
24
15
1996/97
22
18
2003/04
20
15
Source:
Alailima and Mohideen 1983; estimations by authors and Aparnaa Somanathan of IHP.
quintile received 20 percent of government health
spending; the richest quintile, 15 percent.
Government outpatient spending is the most pro-
poor (27 percent went to the poorest quintile in
2003/04 versus 11 percent to the richest quintile),
and inpatient spending is more evenly distributed
(18 percent versus 16 percent). Because Sri Lanka
does not means test access to public services, the
main reasons for the pro-poor targeting of gov-
ernment health subsidies are: a dense network of
health facilities that makes government health ser-
vices physically accessible to the poor, lack of user
charges, and the voluntary opting-out of the rich
into the private sector (Rannan-Eliya 2001).
In effect, what Sri Lanka does is guarantee its poor
effective access to free health services, especially
hospital care, and then relies on differentials in
consumer quality in services to persuade the richer
patients to voluntarily opt to use and pay for private
delivery. The role of these consumer differentials
is discussed later.