CHAPTER 3

Economic growth and human development

Economic growth is essential for human effective means of sustained human devel-
development, but to exploit fully the oppor- opment. The Republic of Korea is a stun-
tunities for improved well-being that growth ning example of growth with equity. Sec-
offers, it needs to be properly managed. ond, countries can make significant im-
Some developing countries have been very provements in human development over
successful in managing their growth to long periods - even in the absence of good
improve the human condition, others less growth or good distribution - through
so. There is no automatic link between well-structured social expenditures by gov-
economic growth and human progress. One ernments (Botswana, Malaysia and Sri
of the most pertinent policy issues concerns Lanka). Third, well-structured government
the exact process through which growth social expenditures can also generate fairly
translates, or fails to translate, into human dramatic improvements in a relatively short
development under different development period. This is true not only for countries
conditions. starting from a low level ofhuman develop-
ment but also for those that already have
Typology ofcountry experience moderate human development (Chile and
Costa Rica). Fourth, to maintain human
The human development experience in development during recessions and natural
various countries during the last three dec- disasters, targetted interventions may be
ades reveals three broad categories of per- necessary (Botswana, Chile, Zimbabwe and
formance. First are countries that sus- the Republic of Korea in 1979-80). Fifth,
tained their success in human development, growth is crucial for sustaining progress in
sometimes achieved very rapidly, sometimes human development in the long run, other-
more gradually. Second are countries that wise human progress may be disrupted
had their initial success slow down signifi- (Chile, Colombia, Jamaica, Kenya and
cantly or sometimes even reverse. Third are Zimbabwe). Sixth, despite rapid periods of
countries that had good economic growth GNP growth, human development may not
but did not translate it into human develop- improve significantly if the distribution of
ment. From these country experiences income is bad and ifsocial expenditures are
emerges the following typology: low (Nigeria and Pakistan) or appropriated
• Sustained human development, as in by those who are better off (Brazil). Finally,
Botswana, Costa Rica, the Republic ofKorea, while some countries show considerable
Malaysia and Sri Lanka. progress in certain aspects ofhuman devel-
• Dz"sruptedhuman development, as in Chile, opment (particularly in education, health
China, Colombia, Jamaica, Kenya and and nutrition), this should not be inter-
Zimbabwe. preted as broad human progress in all fields,
• Missed opportunitiesfor human develop- especially when we focus on the question of
ment, as in Brazil, Nigeria and Pakistan. democratic freedoms.
The analysis ofthese country cases leads The main policy conclusion is that eco-
to several important conclusions. First, nomic growth, ifit is to enrich human devel-
growth accompanied by an equitable distri- opment, requires effective policy manage-
bution of income appears to be the most ment. Conversely, ifhuman development is

42 ECONOMIC GROWTH AND HUMAN DEVELOPMENT

to be durable, it must be continuously tion, potable water and other social services
nourished by economic growth. Excessive - usually provided by government - and
emphasis on either economic growth or can be measured by the shares of govern-
human development will lead to develop- ment budgetary expenditures in GNP or
mental imbalances that, in due course, will GDP. The level of meso policies can be
hamper further progress. described as low if government expendi-
tures on the social sectors are less than 6%
Policiesfor human development of GDP, moderate if they are between 6%
and 10% and high if they are greater than
Many factors influence the levels and 10%. Per capita public spending in the social
changes in human development, ranging sectors would be expected to rise with aver-
from aspects of the macro economy - age per capita GDP. Richercountriesmay
which in turn are affected by developments thus have higher absolute social spending There is no
in the international economy - to micro per capita even if the level of their meso automatic link
factors operating in individual households. interventions, as defined here, is lower. between economic
Also important is at least one set ofinterme- Higher incomes can, therefore, have a posi-
diate, or meso, variables: the level and tive impact on human development not
growth and human
structure ofgovernment expenditures and only through ensuring high primaryincomes progress
government programmes for the social sec- but also by providing larger absolute re-
tors. Meso policies cover the whole range of sources to the government.
fiscal policies, including those directly af- It is also desirable to distinguish differ-
fecting the distribution of income, but the ent types of expenditures in each social
analysis here is confined to social expendi- sector, such as that on primary and tertiary
tures. It can be broadened considerably education and on preventive and curative
through more research, particularly on the health care. Such distinctions describe the
links between the level and structure of structure of expenditures within a particular
government expenditures and the distribu- social sector - and provide greater detail
tion ofincome. than the allocation of the total budget to
The main macroeconomic determinants different social sectors. A distinction be-
ofhurnan development, together determin- tween recurrent and capital expenditures
ing the levels and changes in household can also be made.
income, are initiallevels and growth rates of The literature provides fairly conclusive
income per capita and initial levels and evidence of the association between differ-
trends in the distribution ofincome. ent rates of success in human development
The main instruments of government and the relative importance given to differ-
for directly affecting human development ent types of spending on social sectors. For
levels are: example, spending on primary education
• Across-the-boardmeso policies: those for and preventive health care is likely to lead to
the provision ofpublic goods and services in substantially larger improvements in hu-
a way that does not discriminate among man development than spending on higher
different social groups or regions, such as levels ofeducation and curative health care
universal food subsidy systems, universal - at least at low initial levels of human
primary education programmes and nation- development.
wide immunisation programmes. Meso policies can be well designed or
• Targetted meso policies: those for the less well designed, and their impact de-
provision of public goods and services to all pends on their context. Government poli-
members of particular target groups in the cies for universal primary education and for
society, such as the food stamp programme universal secondary education are across-
for lower-income groups in Sri Lanka or a the-board meso policies. But the former are
supplementary feeding programme that more likely than the latter to be part of a
attempts to cover all malnourished children well-structured package of meso policies
in a country. where primary school enrolment ratios are
Meso policies centre on health, educa- stilllow.

ECONOMIC GROWTH AND HUMAN DEVELOPMENT 43

 Similarly, there are differences in determinants and with differences in the
targetted interventions. If substantial bene- relative roles ofspecific meso policies. These
fits accrue to nondeserving groups or do not differences will become clearer in the fol-
accrue to deserving groups, the interven- lowing discussion of country experiences
tion is poorly designed. The balance be- since 1960.
tween targetted and across-the-board poli-
cies also matters. Targetted interventions Indicators ofcountry performance
may be appropriate only under special cir-
cumstances, such as a temporary recession Any assessment of human development
or extreme crisis, or only in countries that would ideally use a composite measure,
have the administrative capacity to manage such as the human development index
efficient targetting. The circumstances (HDI) that was introduced in chapter 1.
Social spending, should thus define the extent and duration But the HDI, now available for only one
directed towards ofesing targetted inter- ventions to protect point in time, does not yet allow trend
the poor, must or improve human development. analysis. We could also consider several
compensate for Meso policies become important when indicators separately -life expectancy at
people's primary incomes, especially those birth, mortality ofchildren under five years
uneven Income of age, female and male literacy, and nutri-
of the poorest, are insufficient for them to
distribution obtain the goods and services needed to tional status, especially that ofchildren. But
ensure a decent level of human develop- good time series are also rare for many of
ment. Primary incomes are the disposable these indicators.
incomes of households from the normal A third option - the one chosen here
workings of the economy. They often are - is to select an indicator that has fairly
insufficient in countries where incomes are comprehensive time series data and that
generally low: even if the distribution of correlates closely with other indicators of
income is good, few people have primary human development. The under-five mor-
incomes sufficient to ensure adequate tality rate meets both these requirements.
human development. Primary incomes can Extensive empirical evidence suggests that
also be insufficient where higher incomes reductions in the under-five mortality rate
are badly distributed: the incomes of some usually reflect improvements in nutrition-
people may allow even developed country particularly that of pregnant women, in-
living standards, but for many others the fants and children - as well as achieve-
primaryincomes maybe insufficient to meet ments in education, especially female lit-
their basic needs. eracy. Estimates of life expectancy, in
Well-structured meso policies are turn, are strongly influenced by under-five
needed to compensate for the low primary mortality rates, particularly in developing
incomes ofimportant segments ofthe popu- countries.
lation. Where incomes are generally low but The long-run trends in under-five mor-
the distribution is good, well-structured tality rates thus provide a useful indicator of
across-the-board meso policies are likely to changes inhuman development. But these
be appropriate. In countries with higher rates refer primarily to changes on only one
average income and good growth but skewed side of the human development equation
income distribution, some targetted inter- - the formation of human capabilities.
ventions thatfavour the poorer segments of They do not capture the use of human
society may need to supplement the across- capabilities.
the-board policies. But even here, and
especially in the long run, well-structured Sustained human development
across-the-board policies - along with
changes in the growth process - are likely Countries with durable progress in human
to have the greatest payoff. development often started from very differ-
The patterns of human development ent initial conditions in 1960 and have at
described here are linked with differences times followed quite different routes to
in the relative roles of the macro and meso sustain their success.

44 ECONOMIC GROWTH AND HUMAN DEVELOPlvIENT

The Republic ofKorea even temporary disruptions in the flow of
primary incomes.
The Republic ofKorea has achieved human The disruptions came in late 1979 and
development through fast and equitable 1980, when the country suffered negative
growth. For most of the people, primary growth for the firsttime in 20 years. Sparked
incomes have grown enough to enable by external shocks, the recession was also
improvements in the human condition with- attributable to a bad harvest in 1980 and the
out significant government interventions. political instability after the assassination of
Social sector expenditures as a percentage President Park in October 1979. Its causes
of GDP have been relatively low. could, moreover, be traced to less rigorous
Although Korea's economic manage- economic management during the second
ment and the resulting growth and distribu- half of the 1970s when the government,
tion are undoubtedly superior to that of spurred by the easy availability of foreign
most developing countries, its performance credit, embarked on an ambitious pro-
has not been consistently good. For ex- gramme ofinvestment in heavy and chemi-
ample, its income distribution worsened cal industries.
during the 1970s, in part because skilled The programme ensured a continuation
workers in the heavy and chemical indus- of the remarkably high growth rates of the
tries, whose growth was emphasised during 1960s and early 1970s, but it also swelled
this period, earned far more than unskilled the budget deficits, widened the trade gaps,
workers. In addition, income disparities and increased the external debt. The exter-
between urban and rural areas, rather sig- nal shocks at the end of the 1970s were thus
nificant to begin with, increased further greater than they would have been under
during the 1970s. the more prudent and restrained growth
The main reason was the urban bias in strategy followed earlier.
the country's development strategy, par- The government's response in manag-
ticularly the concentration of resources in ing the economy - and in protecting the
the capital city, Seoul. This bias meant that, most vulnerable groups during and after the
despite remarkable growth, the distribution recession - provides useful policy lessons
of income, while better than that in most for human development. First, it embarked
developing countries, left much to be de- on a comprehensive programme of stabili-
sired. Many Koreans were vulnerable to sation, liberalisation and structural adjust-

TABLE 3.1
Under-five mortality and other basic indicators of human development
Under-five Adult literacy (%) Calorie supply
mortality rate Life expectancy as % of
(per 1,000 live births) (years) Female Male req uirements
HDI
Country 1987 1960 1975 1988 1960 1975 1987 1970 1985 1970 1985 1965 1985
Sustained human development
Korea, Rep. 0.903 120 55 33 54 64 70 81 91 94 96 96 122
Malaysia 0.800 106 54 32 54 64 70 48 66 71 81 101 121
Botswana 0.646 174 126 92 46 52 59 44 69 37 73 88 96
Sri Lanka 0.789 113 73 43 62 66 71 69 83 85 91 100 110
Costa Rica 0.916 121 50 22 62 69 75 87 93 88 94 104 124
Disrupted human development
China 0.716 202 71 43 47 65 70 56 82 86 111
Chile 0.931 142 66 26 57 65 72 88 97 90 97 108 106
Jamaica 0.824 88 40 22 63 68 74 97 96 100 116
Colombia 0.801 148 93 68 55 61 65 76 88 79 82 94 110
Kenya 0.481 208 152 113 45 52 59 19 49 44 70 98 92
Zimbabwe 0.576 182 144 113 45 53 59 47 67 63 81 87 89

Missed opportunities
Brazil 0.784 160 116 85 55 61 65 63 76 69 79 100 111
Nigeria 0.322 318 230 174 40 46 51 14 31 35 54 95 90
Pakistan 0.423 277 213 166 43 50 58 11 19 30 40 76 97

ECONOMIC GROwrn. AND HUMAN DEVELOPME T 45

 ment. Second, it introduced new social
FIGURE 3.1
Sustained human development: country profiles programmes and intensified existing ones.
The government deficit was drastically
Average growth rates of GDP per capita
reduced, expansion in money supply was
10
-- curtailed and inflation was brought under
---- --
BOT __

8 --
----------- ------- ...... ... _-
control. Many macroeconomic reforms in
both internal and external markets were
__ -.KOR--- ----
6 ------ carried out in an economy that had gradu-
ally returned to more extensive controls in
4

2
.:::::::::::::.;:~
....... -z::::::: -
the 1970s after the substantial reforms of
the 1960s.
......•..........•.......... In social expenditures, the coverage of
o population by health insurance was increased
Botswana - - - - BOT'---' from a tenth in 1978, a year after its initia-
Costa Rica cos •..•..
-2 tion, to almost a third by 1981 and to almost
Korea. Rep.----- KOR -----
Malaysia - _ MAL _ _ a half by 1985. In addition, a medical
-4
Sri Lanka - - SRI - - assistance programme for the lowest in-
1960-70 1970-80 1980-87 come groups was introduced in 1979.
Members of poor families (depending on
Under-five mortality rate their income and ability to work) were en-
titled to free or subsidised medical care,
350
especially maternal and child care.
In addition, the public works pro-
300
grammes to provide employment to the
250 poor during crisis periods were temporarily
increased during the recession. These pro-
200 grammes provided an estimated 9.4 million

150
------------ -- ...... man-days of employment in 1980 alone.
Direct income transfers were made to those

100
~
KOR
............
••••cos
--'BOT __
-- ... --- _ who were unable to work and take advan-
MAL -_ --. tage ofthese employment opportunities due
~.. SRI ~ _
to infirmity or old age. Moreover, the Live-
50
....... -;;.:.::~:::::.:.:.:::: . lihood Protection Programme, initiated in
o 1961, was expanded to benefit an estimated
1960 65 70 75 80 85 87 2 million people in 1981, through grants of
cereals and through cash for fuel and tuition
Social sector public expenditure, percentage of GDP expenses.
As a consequence of these effective
24 meso interventions, the human develop-
ment levels continued to improve even
20 during the difficult years of 1980 and 1981,
though at a temporarily slower rate. Mean-
16 while, the major changes introduced in
•••••••••••••.. cos . macroeconomic policies restored price sta-
12
... .....•••• bility quickly. While the 1980s have not
... ' ... -.. ..... ,----
--- ........ _,"" ,.
w ........;::::.::·· SRI been easy years, the Korean economy has
8
'----- ,-,
performed extremely well, promoting the
human development of its citizens.
BOT " "
4 ~ ---.KOR------' ,------------- One important lesson from this experi-
-------- ence is that countries with a generally im-
o pressive growth but a less impressive in-
1975 1980 1985 come distribution may require well-struc-
tured meso interventions, particularly

46 ECONOMIC GROwrH AND HUMAN DEVELOPMENT

targetted ones, during briefperiods ofslower expenditures in Malaysia found that this
growth. A second lesson is that to avoid goal was actually being achieved. Central
lasting damage to the human condition, government expenditures per capita for
macroeconomic adjustments are needed to education, health care, agriculture and
restore growth that has fallen off. pensions were highest in the rural areas in
the early 1970s.
Malaysia The distribution of primary incomes
improved significantly after the effects of
Malaysia's experience shows that growth taxes and expenditures were taken into
alone is no guarantee of human develop- account. The ratio ofsecondary incomes-
ment, butit also shows that human develop- primary incomes plus the incidence of
ment is possible even under conditions of a budgetary activities - to primary incomes
To avoid lasting
fairly inequitable distribution ofincome- declined steadily at higher incomes. Forthe
if effective meso policies are in place. lowest 10% of income recipients, the ratio damage to the
Malaysia in 1960 was a middle-income of secondary to primary incomes was 1.5. human condition~
countrywith moderate human development This group thus received an additional 50% macroeconom'lC
and areasonable income distribution, which in "income" from government activities. adjustments are
subsequently deteriorated. The human Each of the lowest four deciles - the bot-
condition nevertheless improved steadily, tom40%-hadratiosofatleast 1.20, while
needed to restore
with the under-five mortality rate dropping for the highest income decile the ratio was growth that has
from 106 in 1960 to 32 in 1988. Other 0.93. fallen off
indicators also confirm that there has been The Malaysian experience thus shows
a steady and sustained improvement in the that steady improvements in human devel-
human condition. opment are possible even in the context of
Malaysia had good growth after 1960, good growth coupled with poor income
though not as spectacular as that of the distribution - if the benefits of meso poli-
Republic ofKorea. Per capita GDP grew by cies can be distributed equitably.
about 3% a year during the 1960s and accel-
erated to about 5% a year during the 1970s. Botswana
Even in the difficult 1980s, Malaysia man-
aged per capita growth of2% a year, but the Botswana also translated the benefits of
fruits of this growth were not equitably growth into human development through
distributed. The Gini coefficient steadily well-structured meso policies. Botswana
increased between the late 1950s and mid- started offas a low-income country with low
1970s-from0.42in 1958, to 0.50 in 1970 human development and an uneven distri-
and to 0.53 in 1976. There has since been bution of income. Botswana's human de-
some improvement, but the distribution of velopment is among the best inMrica, par-
income remains quite inequitable: the Gini ticularly Sub-SaharanMrica. Its under-five
coefficient in 1984 was still 0.48. mortality rate fell from 174 in 1960 to 92 in
Malaysia's steady success in human 1988, still high but it started from a much
development owes much to well-structured higher level. The rate of reduction com-
across-the-board meso policies. Public pares well with other successful countries
spending in the social sector averaged about over the past three decades. This success is
8% of GDP during 1973-81. Thatlevel of also reflected in the remarkable progress in
spending is not as high as that in Sri Lanka, literacy. Moreover, Botswana has suc-
which had lower income and poorer growth, ceeded, unlike most Mrican countries, in
but it also is not as low as that in the protecting the vulnerable groups during the
Republic of Korea, which had higher in- adverse external circumstances ofthe 1980s.
come and better growth. An important factor in Botswana's steady
Malaysia's meso policies were designed and sustained improvement in the human
to benefit all groups in society, with special condition is the exceptionally high growth
emphasis on the rural areas where the poorer since independence. Per capita GDP in-
people live. A detailed study of public creased impressively at about 10% a year

ECONOMIC GROWTH AND HUMAN DEVELOPMENT 47

 during 1965-80 and about 8% a year during developmentto the end ofthe 1970s may be
1980-87, a period when most African coun- attributed largely to growth and amoderate
tries suffered negative growth. The high dose of meso policies, Botswana's success
growth - based largely on minerals, espe- in maintaining the earlier achievements and
cially on rapid growth in the production and protecting the vulnerable groups during the
export of diamonds - continued during drought has been largely due to extensive
the 1980s despite the drought (agriculture meso policies, particularly targetted ones.
accounts for less than 10% of GDP). But The government introduced compre-
since 80% of the population is rural and hensive and substantial drought reliefmea-
relatively poor, the drought threatened ru- sures after 1982, two ofwhich were particu-
ral incomes and reduced the availability of larly important:
food. • A public works programme provided
Data on income distribution are not employment on infrastructural projects,
available for Botswana, but unless the initial covering an estimated 74,000 workers in
distribution was extremely inequitable, the 1985-86 and replacing 37% of the income
high average growth rates are likely to have lost due to crop failure.
been accompanied by some growth in the • Supplementary feeding programmes
income of even the poorer segments of the were launched for primary school children,
population, at least before the drought. children under five (for all in the rural areas
Macro factors are thus likely to have con- and for the most malnourished in urban
tributed to the steady improvement in areas), pregnant and lactating women and
human development since independence. tuberculosis patients. In 1985-86 there
Data on meso policies, not available for were an estimated 680,000 beneficiaries,
years before 1973, indicate a moderate but nearly 60% of Botswana's population.
rising level during 1973-77 and a stable and Funds were also provided to help repair
fairly high level during 1978-86. Public water systems and to transport emergency
expenditures in the social sectors rose from water supplies to drought-stricken areas.
4%ofGDPin 1973 to about 9% in 1977 and Agricultural relief and recovery pro-
for the most part remained between 9% and grammes assisted small farmers in clearing
10% thereafter. land and acquiring inputs, including free
Thus, while improvements in human seeds. Livestock relief and recovery pro-
grammes assisted with vaccinations and feed
BOX 3.1
and provided a guaranteed market for cattle.
Botswana's drought relief The budgetary cost ofthe drought relief
programme, which reached more than 70%
A decentralised, cross-sectoral, and flex- compiles monthly reports on the nutri- of the population, was about $21 million in
ible monitoring network has helped tional status of all children under five at- 1985-86, or 2% of GDP. Foreign donors
Botswana respond quicklywhen drought tending health facilities. It also gets weekly
affects villages or nomadic herders. reports ofrainfall at 250 recording points
contributed an equivalent amount. The
The system, instituted in the wake of and monthly reports on agricultural con- total cost was thus moderate - showing
the 1982 drought, is headed by an Inter- ditions from Botswana's 120 extension that other poor countries could replicate the
ministerial Committee that has the deci- district offices. lrhetechrllcalcorruTilttee programme. Botswana also developed a
sionrnaking power to channel resources makes regular drought assessment tours system of nutritional surveillance and early
quickly to drought-stricken areas. lrhe to confirm and supplement the network's
warning to permit timely identification of
Committee's information base is con- data.
tinually updated by an Early Warning Timely, local-level information is problems and appropriate interventions, a
Technical Committee that monitors rain- combined with top-level policy involve- system also likely to be replicable in other
fall, food supplies and reserves, agricul- ment to assure quick response. When countries (box 3.1).
tural conditions, and the nutritional status health centres reported falling weights Botswana's achievements in human
of children - and makes district-by- among children in 1984, the Interminis- development have clearly been helped by
district recommendations for drought terial CorruTilttee quickly provided sup-
recovery assistance. plementary feeding supplies for under-
the prosperity of the diamond industry and
The technical committee is sup- fives across the country. Further reports its impulse to growth. Butitis also clear that
ported on the ground by the National of malnutrition in the next year led to the meso policies for the provision of basic
Nutritional Surveillance System, which restoration of full drought rations. health and education facilities across the

48 ECONOMIC GROwn I AND HUMAN DEVELOPMENT

board - together with targetted policies to between 15% and 24%oftotal public spend-
meet special needs during the drought - ing in the 1970s. In addition to the system
contributed much as well, particularly in of food subsidies, the strong interventions
protecting vulnerable groups. initiated in education and health before
independence were maintained thereafter.
Sri Lanka All this is reflected in a high proportion of
public expenditures on the social sector in
Sri Lanka's experience can be divided in GDP: about 10% during 1973-78.
two phases: 1960-78 and after1978. Modest In 1979, in the wake of the change in
growth characterised the first phase, with macro policy, the food subsidy programme
per capita GDP rising about 2.2% a year gave way to a food stamp scheme: only
during 1960-70 and about2.5%during 1970- households whose declared incomes were
80. But the distribution of income was fairly less than a specified level received food
good, with the Gini coefficient of house- stamps, which could be used to buy basic
hold income falling from about 0.45 in 1965 foods from designated shops. This change
to 0.35 in 1973. Mter1978,percapitaGDP was primarily intended to reduce the budg-
growth accelerated to more than 3%, but etary burden of government. The share of
the distribution ofincome worsened. Esti- food subsidies in government expenditures
mates of Gini coefficients for 1978 and dropped from 15% in the mid-1970s to
1982 are comparable to those forthe 1950s about 3% in 1984, and the share in GNP
and early 1960s: above 0.45. dropped from about 5% to 1.3%. Overall,
It can thus be said that Sri Lanka shifted social sector expenditures declined from
from a regime of moderate growth with a around 10% of GDP during 1973-78 to
good distribution ofincome (before 1978) around 7% during 1980-85.
to one of better growth with a poorer in- The relative reduction in social sector
come distribution (after 1978). Through- expenditures was countered, however, by a
out, however, the levels of income have
remained relatively low. This has meant BOX 3.2
that, although growth was moderate and the Food stamps miss the target in Sri Lanka
income distribution good, substantial im-
In Sri Lanka, some of the poorest people The effects were quickly discernible.
provements in human development could lack access to their main staple - rice- The national average daily calorie con-
not be achieved exclusively through the despite a food stamp programme in- sumption per capita was virtually the
macro side, and the meso interventions had tended to help them. The main reason is same in 1981-82 as in 1979: just under
to be significant. lack of flexibility in programme design. 2,300 calories. But the per capita calorie
Indeed, Sri Lanka has a long history of First, the shift from an across-the- consumption of the lowest decile fell
board programme to a targeted scheme from 1,335 calories to 1,181, and that of
social sector interventions dating back to was introduced in 1979 in order to en- the second lowest decile from 1,663 calo-
the period before independence. As early as sure that a relatively large share of the ries to 1,558.
1945, the government had extended free benefits from the government's food sub- In contrast, the calorie consumption
medical care to almost every part of the sidies flows to the most deserving groups. of the rich increased, mainly because
country and introduced universal free edu- However, inflation doubled the price of they appropriated a greater share of the
food between 1979 and 1982 - and fruits of accelerated growth to more than
cation up to the university level.
halved the purchasing power of the food compensate for the cutbacks in food
But its best known meso intervention is stamps, since their face value remained rations.
the nearly universal food subsidy introduced unchanged. As a result, the absolute The lesson conveyed by this experi-
in 1942. That system persisted until 1979, amount of real income transferred to the ence is that the effectiveness of policy
with only occasional changes in the eligibil- poor was, at the end, considerably lower measures, especially that oftargeted pro-
ity criteria and the quantities allowed. For than before. grammes, must be subject to continuous
Second, after March 1980, no new monitoring. This holds true, in particu-
example, the proportion of rationed rice to applicants were accepted for the scheme. lar, for policy measures being imple-
total rice consumed exceeded 70% at one This disqualified all new-borns and fami- mented under conditions of rapid socio-
time but declined to about 50% after 1966. lies that subsequently suffered serious economic change - changing consumer
As a proportion of the total calorie intake, income losses. Meanwhile, many higher- or producer prices, flagging or expanding
rationed rice represented about20%in 1970. income households continued to benefit unemployment, and falling or rising wage
from the scheme by underreporting their levels.
The budgetary cost was substantial, varying
mcomes.

ECONOMIC GROwrH AND HUMAN Dl'VELOPMENT 49

 better distribution of its benefits. In 1973 of its significant improvements in human
the middle-income groups benefitted most, development over a relatively short period.
but by 1980 the per capita benefits declined Costa Rica started as a middle-income
with rising incomes, and the poorest 40% of country with an income distribution that
the population derived more benefit than was fairly moderate, at least for Latin
other income groups from government America, and with amoderate level ofhuman
spending. development. During 1960-87 the trends in
The biggest change was in the distribu- the under-five mortality rates reflected
tion of benefits from education. The par- human development approaching that in
ticipation of low-income children in pri- the developed countries. Notable, how-
mary schools improved markedly in this ever, is the substantial improvement in the
period, so a greater share ofthe expenditure 1970s. The under-five mortality rate fell
Promoting faster on primary schooling accrued to these from 121 in 1960t022in 1988, but much of
economic growth at groups. Moreover, the attempt to restrict that reduction came between 1970and 1980,
the expense of food stamps to low-income groups man- dropping from 76in 1970 to 31 in 1980-
equity can damage aged to increase the proportion of benefits a more than 50% reduction in a decade.
accruing to the poor. But the weaknesses of Growth in the 1960s and 1970s was
the invisible bond
the new programme appear to have led to an fairly good, with per capita income increas-
between the people absolute decline in some aspects ofwelfare, ing more than 3% a year, but it turned
and the such as calorie consumption of the poorer moderately negative in the 1980s. The Gini
government segments of the population (box 3.2). coefficient declined from 0.52 in 1961 to
Sri Lanka's experience thus suggests 0.44 in 1971 but then returned to about
that, in a low-income country with a good 0.50 in 1977. Since then, there has again
distribution of income, well-structured been some decline, but even at its lowest
across-the-board meso interventions can point in 1982 the Gini was still about 0.43.
significantly improve human development. Turning to meso policies, social sector
These policies proved vulnerable, however, expenditures expanded in Costa Rica. GDP
to political and economic changes. In prin- was growing impressively, and the share of
ciple, the shift towards targetted interven- government expenditures in GDP was also
tions should have helped sustain improve- increasing, from 18%in 1973 to about 25%
ments the human development, despite a in 1979. So, although the share of the social
worsening of the income distribution. But sectors in total expenditures remained stable,
in practice, replacing across-the-board meso at the high level ofmore than 50%, the share
policies by targetted policies can worsen the in GDP rose from 10% in 1973 to 14% in
position of some vulnerable groups. 1979.
An important lesson from Sri Lanka's Social expenditures were also well struc-
experience is that promoting faster eco- tured. In the 1970s Costa Rica introduced
nomic growth at the expense of equity - major changes in its health strategies to
without effective social safety nets to pro- ensure complete coverage of basic health
tect human development, especially after a services for the entire population. Under
sustained period of good human progress the first national health plan, launched in
- can damage the invisible bond between 1971, public resources for the health sector
the people and the government and lead to were increased, and efforts were made to
considerable social and political turmoil. increase the efficiency of their use. These
The question for governments in a similar programmes fell into two categories.
position is whether, and to what extent, First, the strategy for primary health
budgetary transfers are necessary if free care was to extend the coverage ofbasical1y
markets fail to protect the poor adequately. preventive services to people not previously
served - through the rural health pro-
Costa Rica gramme (begun in 1973) and the commu-
nity health programme (1976). By 1980
The last example ofsustained human devel- water and sanitation services in rural and
opment is particularly interesting because urban areas reached 60% ofthe population.

50 ECONOMIC GROWTI-I AND HUMAN DEVELOPMENT

Immunisation campaigns against measles, China
diphtheria, pertussis and tetanus were
launched, and sanitation activities (for po- A low-income country with good income
table water and sewage disposal) were in- distribution, China dramatically improved
tensified in rural areas. Community partic- its human condition through extensive, well-
ipation in health programmes was also structured, across-the-board meso interven-
encouraged. tions (with some targetting) during a period
Second, medical services were improved ofarguably moderate growth, roughly 1960-
and systematically broadened, mainly by 78. But even with good subsequent growth,
transferring the ministry ofhealth hospitals reductions in the coverage of meso policies
(often poor in resources and frequently led to a stagnation or, by some accounts,
offering deficient services) to the Social even a reversal of these trends. Moreover,
Security System (CCSS). The CCSS doubled China's record is flawed by the absence of
Social expenditures
the number of centres offering outpatient other vital human choices, including politi- must be
services and tripled the amount of physi- cal and economic freedom. restructured to
cian-hours between 1970 and 1980. There China's achievements show up in the benefit the many,
was thus an important restructuring ofhealth under-five mortality rates, reduced from rather than a few
expenditures: the number of hospitals fell 202 in 1960 to 98 in 1970 and more than
from 51 to 37, and the number of out- halved to 43 in 1988. Other indicators tell
patient installations rose from 348 to 1,150. a similar story.
Also by 1980, insurance coverage for illness There is some controversy over whether
reached 78% of the population. All these China has sustained its progress in the 1980s,
programmes paid special attention to re- a period of significantly faster growth in
gions with lower levels of human develop- incomes. The dramatic reductions in the
ment. under-five mortality rates until 1980 seem
The achievements of Costa Rica's pub- to have slowed during the 1980s, even
lic health programmes should not be con- though the rates are still higher than those in
sidered in isolation. Public health had the the industrial countries.
political support of a government highly A recent World Bank study suggests
sensitive to social needs. It also had the that China's earlier progress in improving
economic support of the growth and pros- the health ofits people may have stagnated
perity after 1964. Improvements in educa- somewhat in recent years. For example,
tional attainments were important as well. there are reports of substantial increases in
The proportion of women who completed the prevalence of schistosomiasis in certain
their primary education rose from 17% in regions of China. Although the evidence is
1960 to 65% in 1980, accelerating the de- not conclusive, it appears that China's
cline in infant and child mortality. achievements through the end of the 197Os
Costa Rica shows that assigning a high may have slowed down considerably, if not
priority to social sector expenditures, reversed, on some fronts in recent years.
coupled with well-structured across-the- Widespread literacy and food pro-
board policies, can dramatically improve grammes to help ensure adequate nutrition
the human condition despite only moderate have been important in China, but the de-
growth and a poor distribution ofincome. velopment of an effective health care sys-
tem has contributed most to improving the
Disrupted human development human condition there. The Chinese health
care system has many noteworthy features,
The countries in this category achieved some of them quite innovative (box 3.3).
success in human development, often dra- • It strongly emphasises preventive health
matic success, but could not maintain it. services over curative.
Like the previous group of countries, they • It mobilises people to carry out preven-
differ in their initial conditions and in the tive health campaigns.
speed ofinitial progress before stagnation • It delivers services even to remote rural
or reversals set in. areas.

ECONOMIC GRO'V:'TH AND llUMAN DE\'ELOPMENT 51

 • It consumes a relatively large propor-
FIGURE 3.2
Disrupted human development: country profiles tion of national resources.
China's advances in human develop-
Average growth rates of GDP per capita
ment are also attributable to socioeconomic
Chile - - - - CLE - - - .
10 China ••••••••••• CNA .......... gains in meeting basic needs. China's ap-
Colombia COL •._ ..- •••••••••••••••• proach to ensuring an adequate food supply
8 Jamaica ----- JAM ----- ••,
to its citizens has differed from that in Sri
Kenya - - KEN - - •. CNA·
Lanka. Food security was for many years
6 Zimbabwe _ . - '" _ . - ••• ~._~./ __/ . built into the commune system, and the
4 production brigades gave their members
::..\! ;::::~
rations of basic foods in exchange for work.
2 ~"
----
........ ......... ----_
------------COL
---------- ------ KEN ------..
Communes sold grain or paid taxes on
............. ----CLE.
o production to the state. The state could
.--.--~~.--
.................. --··ZIM ._._.===-:.~ ---------.. then guarantee food security to communi-
-2 ...........- JAM----------- ties that, for some reason, were short of
food and required relief grain.
-4
Recent changes - the adoption of the
1960-70 1970-80 1980-87 household responsibility system in 1979 and
the dismantling of communes in 1982 -
Under-five mortality rate radically altered the situation, with produc-
tion now being liberalised and left more to
350
communities and even individuals.
The recent economic reforms in China
300
have also led to a collapse of the rural
250 cooperative insurance system, removing the
protection against the financial risks of ill
200~ health for the great majority of rural people.
--:-:_--=:- . __ KEN
'CNA liM - . _ .
Those risks can be substantial because the

---::=-_---: ....-. '

150::--.. -'"

.
-.-. Chinese health system recovers a high pro-
portion of its costs: hospitals typically re-
100
---______
'.....::::~ .....
-..:::.-:,:.......... COL--. _ cover about three-quarters of their operat-
---JAM .. __ - - ......==.•••••••••••••••••• ing costs through user fees and drug sales.
---------------~~~~:::~~~;;~~~.;~~~
50
To put this in perspective, the costs per
o hospital admission average $36 for rural
1960 65 70 75 80 85 87 people, even though the annual rural in-
come per capita is less than $100 in many
Social sector public expenditure. percentage of GDP regions. The costs are about twice as high
($75) for urban residents, but most of them
24 are still covered by compulsory, state-sub-
sidised health insurance.
20 The network of barefoot doctors appar-
ently has been another casualty of the re-
16 forms, with rural health care coverage de-
clining, county hospitals and rural clinics in
12 financial distress, and private medical prac-
tice emerging again.
8 ..-..............- China's new "household responsibility
system" reintroduced the concept of eco-
4 nomic incentives for individual productiv-
ity. But the larger role for private and coop-
o erative enterprises, the growth ofpiece work,
1975 1980 1985 and the establishment of liberalised enter-
prise zones - all part of the post-I978

52 ECONOMIC GROwrH AND I IUMAN DEVELOPMENT

reforms - also worsened the distribution 40% in 1973, increased steadily to reach
ofincome across families and regions. 50%in 1979 and average 60% in the 1980s.
Although the events of 1989 may be So, although the 1970s were a period of
another reversal, peasants and workers until slow growth and declining government
recently were encouraged to produce pri- expenditures, social sector expenditures
vately for individual reward. The post-1978 increased marginally through 1978, having
reforms undoubtedly increased the produc- been 14%ofGDPin 1973.
tion incentives, as reflected in accelerated Chile implemented across-the-board
growth, but they appear also to have hurt, policies but targetted its health program-
probably unintentionally, the variables that mes on maternal and child care. In addi-
contribute directly and indirectly to human tion, a screening programme was estab-
development, slowing the earlier rates of lished in conjunction with well-baby clinic
progress. There is no rationale for neglect- check-ups (already reaching almost all in-
ing social development in a period of accel- fants) to detect and treat children suffering
erated economic growth. from malnutrition. This programme proved
highly effective in protecting the most vul-
Chile nerable groups during a period ofeconomic
instability.
Chile also sawits dramatic progres~inhuman Chile's experience shows that some
development falter. Chile started the 1960s human development indicators can improve
as a middle-income country with a moder- dramatically- even during periods ofpoor
ate distribution of income and a moderate growth - if well-structured across-the-
level ofhuman development, and then had board policies are combined with some
its subsequent progress bring it close to
developed country levels. BOX 3.3
Like Costa Rica, Chile dramatically China's health care system
reduced the under-five mortality rate from
142in 1960t026in 1988, with much ofthe Shortly after the revolution China initi- Recent estimates show that while the
reduction coming in the 1970s. But unlike ated campaigns to improve sanitation by number of western-style doctors per
eliminating the "four pests" (rats, flies, 100,000 population in China is two and a
Costa Rica, which sustained its progress in
mosquitoes and bedbugs), to vaccinate half times that in India, the number of
human development on all fronts, Chile against and cure infectious diseases, and village-level health workers was 4.5 times
appears to have been less consistent. to control the vectors of such major that in India.
Its calorie consumption per capita de- endemic disorders as malaria and schis- Extensive health insurance coverage.
clined slightly between the mid-1960s and tosomiasis. During the early 1980s, financing came
mid-1980s, and debate surrounds the trends The keys to success were mobilising in almost equal amounts from three main
the masses, extending services to the sources: private outlays (32%), labour
in general living conditions since the mid-
remotest areas and making health serv- insurance (31%) and state budget expen-
1970s. It has been suggested, for example, ices affordable. ditures (30%), with the residual financed
that the under-five mortality rate has de- Mobilising the masses. The Chinese by production brigades. Notable in this
clined despite a steady deterioration in tackled the "free-rider" problem that financing profile is the high proportion of
overall living standards - reflected in sharp often hampers effective preventive health expenditures mediated through insur-
measures by making people responsible ance schemes, a reflection of extensive
falls in real wages, worsening income distri-
for them. According to some estimates, health insurance coverage. The coverage
bution' rising incidence of certain diseases, preventive measures accounted for less of insurance changed drastically after the
deteriorating housing conditions and falling than 5% of the resources devoted to economic reforms of the early 1980s. In
primary school enrolment ratios. It has also health. Mass mobilisation of surplus 1981 about 70% of the population was
been suggested that poverty declined for labour, especially during slack agricul- completely insured. But there were
the extremely poor but not for the poorer tural seasons, prevented the overtaxing considerable urban-rural differences in
of the health budget in a poor society- health spending. Urban expenditures,
groups as a whole.
but achieved outstanding results. estimated at about $16 per person, were
Chile's growth was moderate in the Extending services to remote rural ar- more than triple the rural expenditures.
1960s, flat in the 1970s and negative in the eas. The deprofessionalisation of health State subsidies for health in urban areas
1980s. Total government expenditures fell care providers through a mass cadre of were almost 10 times those for rural areas
from35%ofGDP in 1973 to 30%in 1980. barefoot doctors at the grassroots helped - about $13 per capita compared with
extend basic services into remote regions. less than $1.50 per capita.
But the share of social sectors in total gov-
ernment expenditures, already very high at

ECONOMIC GRO\Vlli AND I rUM-AN DEVELOPMENT 53

 targetted interventions. It also shows that 3% a year. There was some improvement
the targetted interventions may not help in during the 1980s, but growth rates remained
maintaining overall progress if growth does negative. The effect was harshest on the
not recover. poor, who suffered falling real incomes and
unemployment. The government tried to
Jamaica maintain real wages and protect the vulner-
able groups, with only partial success,
Like Chile, Jamaica started the 1960s as a through food subsidies, price controls and
middle-income country with a moderate employment schemes. But with growing
income distribution and a moderate level of external and internal deficits - the current
human development. But instead ofhaving account deficit was more than $200 million
dramatic improvements in the human con- and the government's budget deficit more
Targetted social
dition, its progress was more uniform. And than 15% of GDP in 1981 - these meso
spending may not it had even less success than Chile in sus- policies were difficultto sustainindefinitely.
help in maintaining taining its progress during the 1980s. For Mer the change in government in 1980,
overall progress if example, the number of children admitted government expenditures, including those
growth does not for malnutrition to the country's major on the social sectors, were cut as part of an
children's hospital more than doubled be- adjustment programme, reducing the share
recover tween 1978 and 1985. ofsocial sector public expenditures in GDP.
Jamaica's growth rates were very re- With per capita real GDP falling during the
spectablein the 1950s, with per capita GDP 1980s, real per capita expenditures on the
increasing by nearly 7% a year on average, social sectors fell as well. Education expen-
and still reasonable in the 1960s, with per ditures per person under 15 are estimated
capita GDP growing about 3.5% a year. to have declined 40% and per capita health
Data on income distribution suggest, how- expenditures by 33% between 1982 and
ever, that a very inequitable distribution 1986.
became even worse. The proportion of The government's attempts at targetted
aggregate income going to the poorest 40% interventions achieved limited success. It
of the population, only 8.2%in 1958, fell to introduced a food aid programme in 1984
7% in 1972, while the share of the richest to protect the most vulnerable - infants,
10% ofthe population increased from 43.5% school children, pregnant and nursing
t050%. women, the elderly and the very poor people,
No information is available regarding together constituting about half the pop-
public expenditures in the social sectors ulation. But the per capita benefits from
before 1976, but data for 1976 onwards the scheme fell short of the requirements.
show a high level of meso interventions in The adverse movements in the macro and
the social sectors, particularly for 1976-80. meso determinants ofhuman development
Public expenditures in the social sectors slowed and in some cases reversed the rate
ranged between 12% and 14% of GDP of progress.
during 1976-80 and then fell to about 10%
in 1985 and 1986. Ifsocialsectorexpendi- Colombia
tures in the 1960s and early 1970s were
comparable to those subsequently, the meso Per capita GDP in Colombia grew moder-
interventions before 1976 were quite im- ately at 2.1% a year in the 1960s and 3.7% a
portant. It is thus likely that the steady year in the 1970s. The country did avoid a
improvement in human development dur- recession in the difficult 1980s, but its per
ing the 1970s may have been facilitated by capita GDP growth nevertheless slowed to
respectable growth, complemented by meso about 1% a year.
policies that compensated somewhat for Although modest, Colombia's economic
the skewed and worsening distribution of growth in the 1980s made it possible for the
income. government 10 maintain the per capita in-
Growth deteriorated, however, during creases in social expenditures. The share of
the 1970s, with per capita GDP falling about public spending on education in GNP

54 ECONOMIC GROWTH AND HUMAN DEVELOPMENT

moved up from 1.7% in 1960 to 2.8% in ceived 46%, representing 25 times the in-
1980, and that on health from 0.4% to 0.8% come of the poorest 10% of households.
of GNP. But the adjustment programme The government's policies only partly
adopted in 1984 reduced public expendi- offset the effects ofbad distribution. Kenya's
tures, including the expenditures on social meso interventions have generally been
sectors. Social spending nevertheless con- moderate, with the share of social sector
tinues to account for about one-third of public expenditures remaining remarkably
total public expenditures. stable at 7%to 8% ofGDP during 1973-86.
Colombia's human development indi- Two-thirds of this spending was for educa-
cators mirror the overall economic situ- tion' and the rest mostly for health. Kenya's
ation. The country's income distribution education system also benefitted from vol-
improved in the 1970s and 1980s, with the untary self-help (harambee) efforts. In 1970,
Gini coefficient declining from 0.57 in 1971 for example, two secondary students in five
to 0.45 in 1988. The under-five mortality were in unaided (mainly harambee) secon-
rate fell from 148 in 1960 to 78 in 1980. dary schools.
Since then, the decline has been more In Kenya, therefore, the government's
modest - to 68 in 1988. There has also moderate efforts were supplemented by
been a slowdown in the growth ofreal wages significant private involvement in the provi-
since 1987, but this decline appears to have sion of social services, especially in educa-
come to a halt. tion. This, along with a moderately good
Although the slower growth in the 1980s growth, contributed to Kenya's improve-
would have required a compensatory in- ments in human development through the
crease in meso interventions, some elements end of the 1970s. In the 1980s, however,
of the government's earlier policy package the failure to increase the coverage ofmeso
-like its effective food stamp scheme - policies in the face of declining primary
were discontinued, primarily for fiscalrea- incomes and unequal income distribution
sons. The economic adjustment policies appears to be associated with a deteriora-
appear, however, only to have slowed human tion in human development.
progress, not to have reversed it. The chal-
lenge now is to convert the gains in eco- Zimbabwe
nomic growth that are expected from these
policies into further improvements in hu- Human development in Zimbabwe, rela-
man development. tive to the rest of Sub-Saharan Mrica, has
been very good. But Zimbabwe also suf-
Kenya fered some stagnation after progressingfrom
poor initiallevels of human development.
Human development in Kenya was for many Improvements in Zimbabwe have come
years successful, despite difficult initial despite steadily worsening growth since the
conditions -low income, low human de- 1960s, with per capita GDP falling about
velopment indicators and a rather uneven 1%a year in the 1970s and about 1.5% a year
distribution of income. But the progress in the 1980s. Data on income distribution
has slowed down in recent years. are being collected only now, but it is widely
A low-income country, Kenya had a agreed that inequalities were significant
reasonably good growth in the 1960s and in before independence and have since been
the 1970s, when its per capita GDP in- reduced by the redistributive policies of
creased at about 3% a year. But like most government - but only somewhat, leaving
African countries, it suffered negative growth substantial inequality.
in the 1980s, with per capita GDP falling Zimbabwe's improvements in human
about 0.9% a year. Detailed data on income development can therefore be attributed to
are not available, but one estimate suggests social sector expenditures, which were
an inequitable distribution. In 1976 the moderate to high before independence in
poorest 40% of households received only 1980. The country's experience since inde-
9% of total income while the top 10% re- pendence shows the difficulties of sustain-

[CO 'O\IIlGRO\X'11IA DIIl'.\lANDLVELOl',\lENT 55

 ing human development, even with well- prevented a reversal during the prolonged
structured meso policies, if growth remains recession (box 2.4). Zimbabwe's experi-
negative for long periods. ence shows that although it may not be
After independence, the government possible to rely on meso policies alone to
gave greater prominence to social sector sustain progress in the face of poor growth,
meso policies and restructured its social improvement in their structure can avoid
spending towards activities having a greater reversals, at least in the short run.
impact on human development, targetting
those in need. These expenditures jumped Missed opportunities
to more than 10% of GDP after 1980. for human development
At the time ofindependence, Zimbabwe
inherited a highly inequitable health care Brazil
Reversals in human system - reflected, for example, in the fact
development that 44% ofpublicly funded services went to Brazil failed to achieve satisfactory human
during adjustment sophisticated central hospitals that served development despite high incomes, rapid
periods can be only 15% of the population, while only 24% growth and substantial government spend-
went to rural health services for the majority ing on the social sectors.
avoided through
of the population. After independence, An upper-rniddle-income country, Bra-
careful policy several measures were taken to redress these zilhad a per capita GNP of$2,020 in 1987.
management imbalances. Except for 1980-87, when its per capita
• Health care became free for those earn- GDP grew at just over 1% a year, Brazil's
ingless thanZ$150 a month, the vast major- growth has been quite good - with average
ity of the population. annual growth of per capita GDP hovering
• The programme ofimmunisation against around 3% in the 1950s and 1960s and
six major childhood infectious diseases and rising to a very respectable 6.4% in the
tetanus immunisation of pregnant women 1970s.
was expanded. The proportion of fully Central (federal) government expendi-
immunised children between 12 and 23 tures in the social sectors ranged between
months is estimated to have increased from 8% and 10% ofGDP during 1973-86. As a
25% in 1982 to 42% in 1984 in rural Zim- percentage of total central expenditures,
babwe and from 48% in 1982 to 80% in they remained at about 50% during 1973-
1986 in Harare City. 79 but fell to35%in 1986. Surprisingly, the
• A programme for building hospitals and level ofmeso policy interventions was quite
rural health centres was initiated - con- high, even in comparison with countries at
structing 163 rural health centres byJanuary comparable income levels. Moreover, the
1985 and upgrading numerous rural clinics social spending by state and local govern-
and provincial hospitals. ments matched that of the central govern-
• A diarrhoeal control programme was ment. Total social expenditures by all levels
launched in 1982, and a Department of of government and by the private sector are
National Nutrition was established - for estimated to have constituted a quarter of
nutrition and health education, for growth GDPin 1986.
monitoring and nutrition surveillance and Despite rapid growth and substantial
for supplementary feeding programmes for meso interventions, Brazil's human devel-
children. opmentrecordhas been unsatisfactory. The
In addition to these measures, most of under-five mortality rate was still 85 per
which meet the criteria for well-structured 1,000 in 1988, almost twice Sri Lanka's and
meso policies in the health sector, there has only slightly lower than Myanmar's, coun-
been similar restructuring of education to tries with per capita incomes amounting, re-
increase the share of primary education in spectively, to a fifth and a tenth of Brazil's.
total public spending. Life expectancy was 65 years in 1987, and
Although these improvements were not the male and female literacy rates respec-
enough to prevent a slowdown of progress tivelywere 79% and 76% in 1985.
in human development, they may have These national averages hide significant

56 ECONOMIC GROWTH AND HUMAN DEVELOPMENT

regional differences. In the poorer North-
FIGURE 3.3
east, for example, infant mortality rates were Missed opportunities for human development: country profiles
more than twice those in the rest of Brazil in
Average growth rates of GDP per capita
1986 (116 compared with 52), life expec-
tancy at birth in 1978 was only 49 years 10
compared with 64 in the rest of Brazil, and
8
child malnutrition was twice the national
average. 6

............ _.... --
BRA· ............ - - ..... , .....
There are two important reasons for .............
.....
such poor human development in Brazil.
One is the extreme inequality of income
4

2
._:;:J....... -....
- -~PAK
.------
.--.------~ ............
.--",
.."".,.", .. , ..........

...---
......_

distribution. The other is the inefficient

targetting ofpublic resources. The distribu-
tion of income in Brazil is among the worst
o --' NIG' "'

"'.'
......

in the world, with the Gini coefficient esti- -2 Brazil ----BRA - - - .

Nigeria -·-NIG _ . -
mated at 0.60 in 1976,0.56 in 1978,0.56 in Pakistan ..........··...... PAK ........- ......
-4
1980 and 0.57 in 1983.
As indicated earlier, well-structured 1960-70 1970-80
meso policies can compensate for a poor
distribution of income and improve the Under-five mortality rate
human condition. This has not happened in
350
Brazil because public resources did not reach
the poor or improve the basic dimensions of
human development. Substantial public
300
-'-. -.~.
~''''''''
subsidies were provided for "private" goods, 250 "~"",~,~,"_ ~._.::-.""",. ~
usually consumed by the better-off sections
..........--..:~·NIG.
of society, while "public" goods and services 200 --..-.... ..---...
likely to have the widest impact on human ---:"::::::::.::
welfare were neglected. 150 ----- -----SRA
Brazil spends large amounts on social ---------
security (7.4% of GDP in 1986) and on 100 ------------- ----.
housing (2.9% of GDP), with the benefits
50
going disproportionately to the urban em-
ployed. Expenditures on social security o
may not have increased inequality since 1960 65 70 75 80 85 87
they are financed mainly by the beneficiar-
ies, but a considerable amount ofthe expen- Social sector public expenditure, percentage of GDP
diture in housing is for subsidies. Spending
on health and nutrition, by contrast, had a 24
lower priority: about 2.2% of GDP went for
health at all levels of government in 1986. 20
In health, preventive programmes -
such as immunisation, prenatal care and 16
vector-borne disease control - are esti-
mated to be about five times more cost- 12
effective than curative programmes in re-
ducing mortality. But an estimated 78% of
____,' .... - --·SRA-------- . . . . . . ' ",~~- -- --
8
all public spending on health goes to largely
_'-NIG
curative, high-cost hospital care, mainly in 4
urban areas and especially in the urban
..- - - -_ _- - - - - - P A K - - ·
South. This is in sharp contrast with the o
87% of public health expenditure that Bra- 1975 1980 1985
zil allocated to preventive care in 1949, a

ECONO:-"llC GRO\\:'TH A D lIU,\1AN m.\TLOP.\lE 'T 57

 share that fell steadily to 41%in 1961 and to equal between 1960 and 1980, with the
a low of 15% in 1982 before rising to 22% in Gini coefficient for the late 197Os reported
1986. to be about 0.60.
Similarly, more than a quarter of all N or have the supplies of the goods and
public spending on education went to higher services that contribute to human develop-
education in 1983, and only half to primary ment been adequate. The availability of
education. Total public spending per stu- food, for example, is estimated to have
dent in higher education, where the benefits fallen by nearly a quarter between 1965 and
accrue overwhelmingly to higher-income 1975. The accompanying sharp rises in
groups, was about 18 times that in secon- food prices suggest that food supplies did
daryandprimaryeducation. AWorldBank not keep pace with demand.
study shows that 13% of all children in Detailed time series on the level and
ubstantial111eso Brazil come from households receiving less structure ofsocial sector public expenditure
policy than one minimum salary, but they account are not available, and the IMF provides data
il1te ventions, if for only 1% ofhigher education enrolment. only for some scattered years. But other
poorly st1'tlctured Children from households earning more evidence on per capita public expenditures
than 10 times the minimum salary account in the health sector for 1964, 1970and 1976
and badly
for 48% of the enrolment but constitute show very low levels both in absolute terms
tmgetted. can1tot only 11 % ofall children in the country. That and in comparison with countries at similar
make up for a1t is not the only inequity in the system. Spend- incomes. For example, in 1976, total ex-
tmequc I ing per pupil is lower in municipal than in penditure (current plus capital) was only
dist1 ibutl HZ of state schools, lower in rural than in urban about $1.75 per capita. By contrast, in 29
schools and lower in schools in the North- countries with a GNP per capita between
l1lcome east than elsewhere. $300 and $599, the per capita government
Brazil thus demonstrates that substan- expenditure on health exceeded $2 in 18
tialmeso policy interventions, if poorly struc- countries and $6 in 11 countries.
tured and badly targetted, cannot make up The bias in public spending towards
for an unequal distribution of income - curative services was also heavy. For ex-
even if the overall growth ofincome is more ample, in the second five-year plan (1970-
than adequate. 74),80% offederal capital expenditure was
earmarked for teaching hospitals and urban
Nigeria areas. Lagos, with about 4% of the popula-
tion in 1970, had more than 90% of all
Nigeria's moderate rates of growth did not registered medical practitioners in 1973,
lead to substantial progress in human devel- 67% of all state hospitals and clinics and
opment. Its per capita GDPincreased only 72% of all private clinics. This strong bias
0.6% a year in the 1960s, partly as a result of towards curative care in urban areas meant
the civil war. The discovery ofoil led to per that only a small proportion of the rural
capita GDP growth ofa very respectable 4% population had access to medical services.
a year in the 1970s. In 1980 its per capita One estimate suggests that only 25% of
GDP ofabout $1,000 was one ofthehighest Nigerians, most of them in urban areas, had
in Mrica, classifying it as a middle-income health coverage in 1975.
country. This trend reversed in the severe Education received a higher priority than
recessionofthe 1980s, with per capita GDP health in the national plans. In 1977, for
falling about 5% a year during 1980-87. example, education absorbed more than
The unsatisfactory progress in human 40% of the recurrent federal budget and
development, despite rapid growth in the 55% of the recurrent state budgets, but
1970s, can be attributed to several factors. these figures conceal the neglect of primary
The fruits ofrapid growth do not appear education. Although universal primary
to have been distributed equitably. Evi- education was a major objective in the mid-
dence on income distribution is weak and 1970s, the structure of the government's
scattered, but there is general agreement education spending has not reflected this
that the distribution was getting more un- priority. Primary education received less

58 f: (ONOMIC GRo\X'TH AND HUMAN DEVELOPMENT

than 20% of public current educational literacy rate in 1985 was strikingly low at
expenditure in 1981, among the lowest ra- 30%, with large gender disparities - female
tios in Africa. literacy was 19%, male literacy 40%. And its
A systematic analysis of the distribution gross primary school enrolment ratio was
of the benefits of public expenditure in still only 40% in 1987. Pakistan is still far
1977-78 concluded that the federal from universal primary education, some-
government's capital expenditure was un- thing Sri Lanka has already achieved and
ambiguously pro-rich in both urban and China is pursuing. Again, the gender dis-
rural sectors, although the distributional parities are wide: in 1987 fewer than a third
incidence offederal recurrent expenditure of Pakistani girls were enrolled in primary
among urban and rural households was schools, compared with half the boys.
rather proportional and tended to maintain This dismal performance despite re-
the status quo of income distribution. At spectable rates of growth and a moderate
the upper end of the income distribution, income distribution can be explained by the
however, there was a tendency for benefits failure of meso policies. Although growth
to rise as a proportion of income. So, the has been good, Pakistan is still alow-income I
structure of public expenditure in Nigeria country. This low income implies that pri-
did nothing to compensate for the maldis- mary incomes - even if equitably distrib-
tribution ofincome. uted, which they are not - are insufficient,
Nigeria thus provides a clear example of on their own, to permit the bulk of the
failed trickle-down-ofmissed opportuni- people to acquire the goods and services
ties for human development. Rapid growth needed for a decent life. Pakistan thus
did not significantly improve the human needs well-structured meso policies to pro-
condition because of basic flaws in the mote human development - policies that
growth process and the failure to restruc- have been gravely deficient.
ture meso policies to compensate for them. Several factors explain the failure of
rapid economic growth to translate itself
Pakistan into satisfactory human development.
Education and health are provincial respon-
Pakistan's GDP per capita rose almost 4% sibilities, but the provinces lack adequate
a year in the 1960s. Although the rate of financial resources - and a major decen-
growth dropped to 1.6% a year during the tralisation of financial powers from the
1970s, it became respectable once again federal government to the provincial gov-
during the 1980s, increasing at about 3.5% ernments in accord with the 1973
a year during 1980-87. The distribution of constitution has been pending with the
income has been moderate. National Finance Commission since 1974.
But the country's human development There is also a serious imbalance between
has been unsatisfactory, particularly when military and social expenditures - an
contrasted with that of Sri Lanka, whose imbalance that grew much worse in the
growth before the 1980s was fairly modest 1980s as military expenditures rose five
and whose per capita income has been times while public sector development
broadly similar. In 1987, life expectancy in expenditures only doubled.
Pakistan was only58years, much lower than Pakistan spends a very small part of its
Sri Lanka's 71 years and even below the budget on the social sectors - and a large
average of 61 years for low-income coun- and growing part on the military, preempting
tries, among which Pakistan counts as one scarce resources that could otherwise be
of the richest. Likewise, its under-five earmarked for education and health. Only
mortality rate - 277 in 1960, compared 2.2% ofPakistan's GNP went for education
with 202 in China -was still 166 in 1988, and health in 1986, compared with 6.7% for
compared with 43 in China. military expenditures. Military spending
Pakistan's performance on other im- was three times the spending on education
portant basic indicators ofhuman develop- and health. Even adding the fairly consid-
ment leaves much to be desired. Its adult erable spending by the provincial govern-

ECONOMIC lrRO\\'11 I Il 59
 ments on education and health does not nationwide immunisation programme was
challenge the overall conclusion: Pakistan financed by postponing the construction of
spends too little of its GNP on social devel- an expensive urban hospital. Education
opment. spending was tripled in the last four years.
Moreover, a large part of the limited And a special tax was levied on all imports to
social expenditures goes to lower-priority finance additional spending on education.
activities. Of the public current expendi- Pakistan's overall experience shows that
tures on education, 24% was for tertiary inadequate social spending and poorly struc-
education in 1985, compared with 7% in Sri tured meso policies can prevent a low-in-
Lanka in 1986, and only 40% was for pri- come country from improving the human
mary education. There appears to be a condition even if there is rapid economic
similar bias towards lower-priority activities growth with a relatively moderate distribu-
in health spending, but some recent policy tion of income.
changes are steps in the right direction. A

60 ECONOMIC GROWn I AND HUMAN DEVELOPMENT