The Impact of the

Green Revolution and

Prospects for the (Future

Per Pinstrup-Andersen
Peter B . R.. Hazell

Reprinted from
Food Reviews International
VOl 1, No.1 (1985)
 Food Reviews International, 1(1), 1-25 (1985)

THE IMPACT OF THE GREEN REVOiLUTION

AND PROSPECTS FOR Tl-iE FUTURE

PER PhVSTRUP-ANDERSEN
PETER B. R. HAZELL
International Food Policy Research Institute
Washington, D.C.

The Green Revolution (a term used for rapid increases in wheat and rice
yields in developing countries brought about by improved varieties combined
with the expanded use of fertilizers and other chemical inputs) has had a
dramatic impact on incomes and food supplies in many developing countries.
However, the impact goes far beyond these immediate and very important
results. The Green Revolution has facilitated institutional and social changes in
rural areas and provided opportunities for self-sustaining economic growth and
reduced poverty.

The fmal outcome of technological change is influenced by the institutional

and policy environments within which it is introduced. Where such environ-
ments have been favorable, the distribution of the benefits has been wide-
spread, but where they have been unfavorable and appropriate changes have
not been made, the potential benefits irom the Green Revolution to promote

This article has benefitted from comments by Joel; Anderson, Robert Herdt, Peter Jennings, John
Mellor, lohn Pino, and James Ryan on an earlier draft.

Copyright © 1985 by Marcel Dekker, Inc. 0755-9129/85/0101-000 I S3.50/0

2

economic growth. reduce poverty. and facilitate self-sustaining def/clopment

have not been fully exploited. TIle importance of the interaction between the
Green Revolution and the illstitulional and policy environments has been
recognized in much of the debate about the inlpact of the Green Revolution.
However, in spite of such recognition, a great deal of the literature has failed to
distinguish between the impact of the two. This has often lead to the argument
that the Green Revolution is to blame for undesinlble developments occurring
as a consequence of inappropnate~ institutions and policies. Technological
change offers an opportunity for enlarging total agricultural production
through mnre productive use of resources. \Vhethcr the opportunity is used
and how the, associated benefits an~ distributed among population groups ~re
largely a question of institutions and public polky although, as further dis-
cussed below, the nature of the technology plays ill role as well. This does not
mean thaI technological change should await appropriate institutional change.
The fonncr may help bring about the latter and both should be pursued
simultaneously.
This article explores these and r(~lated issues. As the Green Revolution has
matured. more has been learned about its actual and potential impacts. Errors
of judgment and incorrect predictions based on earlier studies are gradually
b"ing conected" Initial st tidies foclLlsed on immediate effects and onen er-
roneously made long·term predictions on the ba;;;is of data from the initial
phases of the Grecn Revolution. ~'fiOre recent w0rk has benefittfd from data
fronl a longer period, thus providing a better basis for predictions and policy
recommendations. Longer-term implications such as the multiplier and en-
vironmental effects and future organization of the agricultural research that
makes continued lec 1ul0logical change possible are also receiving more aUen-
tio'} than previously.
This article addresses some of thie issues considered to be of great impor-
tance for continued .mccess of the Green Revolution. The most recent cata
on its iJnpact on food productkm afte discussed first. Then follows a discussion
of the impact on production fluctuations. Current evidence of the impact on
poverty and nutrition :s summarized in the third section. Recent research has
sho ,In that the multiplier or linkagl~ effects of technological change may be
very important for assuring a desirable path of self-sustaining growth. This issue
is dealt with in the fourth section. followed by (Jl discussion of the role of
\f,yomen in tcchnologicai change, an important growth as well as equity issue
\lJhich has received little attention until recently. Then follows a brief assess-
ment of the actual and potential environmental effects. A number of other
issues with important implications for the future contribution of technological
change, such as future control over genn plasm. organization of and control
over the international agricultural research institutes, the needs for institu·
tional and policy changes in many developing countries, and the need for new
GREEN REVOLUTION 3

technology to facilitate a solution to the acute food probtemin Sub-Saharan

Africa, are discussed in the last section.

mtPACr ON FOOD PRODUCfION

The impact of the Green Revolution on wheat and rice production is a functJon
of the area sown to the new wheat and rice varieties, and the increase in yields
per unit of land. Increasing yields have made rice and wheat more profitable
for farroers than certain other crops, Thus, in addition to yield increases on
traditional wheat and rice land, more land has been brought into cultivation of
these two crops. The Green Revolution has also facilitated significant expan-
sion of irrigation and nlultiple cropping in many countries, thereby adding to
the total ac:eage of these crops, Shorter growing periods and reduced photo-
periodicity are important properties of the new varieties which have enabled
increased multiple cropping,
It is estimated that between one-third and one-half of the rice areas in the
developing world is grown with high-yielding varieties, Table 1 shows estimates
for 11 Asian countries, varying from 9% in Thailand to 78% in the Philippines,
In Latin America, about 2 million ha of rice is irrigated and another 1 mil-
lion grown under favorable rainfalJ and soil conditions but without irrigation
(l), AcclJrcing to estimates by Dr. Peter Jennings, head of CIAT's I rice pre"
gram, high-yielding varieties ~re now grown on about 90% of these 3 million ha.
Estimated yield increases due to these varieties are about 1 ton/ha in irrigated
areas and 0.75 tons/ha on favored upland rice areas, Thus, total annual

I International Center for Tropical Agriculture, Colombia,

Table 1, Area Grown with Modern Rice Varieties in 11 Asian Countries

Country Year 1000 ha % of Rice Area

Bangladesh 1981 2,325 11

'::'k

India 1980 18,495 47

Nepal 1981 326 26
Pak.istan 1978 1,015 50
Sri Lanka 1980 612 71
Burma 1980 1,502 29
Indonesia 1980 5,416 60
Malaysia W 1977 316 44
Philippines 1980 2,710 78
Thailand 1979 800 9
South Korea 1981 321 26

Source: Ref. 2,
4

production increases in Latin America are around 2.5 million tons. Assuming
a price of 5200/ton of paddy ricc, the value of the production increase for I
year is $500 million. Although these estimates are rough and subject to con-
siderable error, it is clear that the production impact is very large.
Such conclusion is supported with c~vidence from Asia. Herdt and Capule (2)
estimate that modern rice varieties added 27 million tons to the production of
rice in eight Asian countries (Bumla, Bangladesh, China, India, Indonesia.
Philippines, Sri Lanka, and TIlailand) which produced 85% of Asia's rice in
1980. Another 29 million tons was added by fertilizers, and irrigation con-
tributed 34 miHion tons. It should be noted that since modern varieties general-
ly have a higher response to fertilizers and possibly other inputs, a clear
distinction among the contribution of each factor is difficult.
Earlier estimates of the production impact of modern rice varieties were
considerably below those reported above. The impact wa~ estimated to be
about 10 million tons in the Far East and less than one-half of a million tons in
Latin America for the year 1976/1977 (3). While differences in estimation'pro-
cedures explain some of the variation between the two sets of estimates, rapid
increases in adoption since 1976/1977 account for a large part, particularly in
Latin AInerica. Rapid increases in rice production due to modern varieties and
associated inputs were not limited to the late sixties and early seventies but are
still going on. In fact, except for Colombia, the bulk of the increase in Latin
America has occurred since the mid-seventies.
The wheat areas grown with mod'ern varieties arc of magnitudes similar to
those for rice. It was estimated that about 30 million ha were grown 'vith these
varieties during 1976/1977 (3) and about 35 million today (5). However, since
much morc land is used for rice than for wheat in developing countries, modern
varieties occupy a larger percentage of the wheat area. James (5) estimates that
the contribution of the new varieties to increased wheat production in develop-
ing countries was 7 million tons in 1982{1983 worth $ 1200 million. Earlier
estimates for 1976/1977 were 19.9-26.7 million tons (6) and around 21 mil-
lion tons worth $2500 million (3). The large difference among the estimates is
due primarily to the different assumptions about average yield increases which
are not known with a high degree of precision. However, irrespective of
whether the actual increase is cioser to the lower or the higher estimates, it is
still extremely impressive.
\Vhile the tenn Green RelJolu tion was originally used for wheat and rice,
high-yielding varieties have been developed for a number of other food crops
important to the developing countries. These include sorghum, maize, cassava,
and beans. The area grown with improved maize varietie~ and hybrids derived
from CIMMYT 2 germ plasm in developing countries runs into millions of
hectares (5). Massive efforts to develop high-yielding technologies for nlany
other food crops grown under developing country conditions are of a more

1 International Center for Maize and Wheat, Mexico.

GREEN REVOLUTION

recent date, and attempts to estimate the global production impact would be
prernature. However, evidence for some crops in a few countries, e,g., beam)
and cassava in Cuba and beans in various Central American countries, shows
considerable promise. SimiJarly, results from farm trials of improved varieties
and cultural practices for various food crops show a great potential for yield
Increases.

~tPAcr ON PRODUCTION FLUcrUATIONS

As shown above, the Green Revolution has enabled many developing countries
to achieve ilnpre ishc-. rates of growth in national foodgrain production since
the mid-1960s.'"\t till: same time though, dlC variability of national foodgrain
production around H.e trend has also increased. India, for example, increased
its average cereal production by 47% between the periods 1952/1953-1964/
1965 and 1967{1968 -1977/1978. At the same time the coefficient of variation
around trend of total cereal production increased from 4.7% in the first period
to 5.9% in the second period (7).
Despite this increased variability, countries like India are still much better
off, even in drought years, in ensuring national food consumption because of
the increased food output these technologies have permitted. But increased
production variability can, in the absence of stabilization policies, lead to more
volatile prices, creating problems for farmers and poor consumers alike. The
degree of price instability induced can be quite large in countries where a high
proportion of total production is consumed on the fann. This is because year-
to-year fluctuations in production are then transmitted to relatively thin
markets. In India, only one-third of foodgrain production is actually marketed,
and farm pIice variability has more than doubled since the mid-1960s for wheat
and rice.·
Mama (8) argues thnt much of tltis increased instability is due to the wide-
spread adoption of improved seed/fertilizer-intensive technologies since the
mid-l 960s. Similar arguments are made by Barker et al. (9). The yields of crops
grown with the new technologies may be more sensitive to weather and disease.
Perhaps more importantly, because they require modern inputs, such as
fertilizer, their yields may also be sensitive to year-to-year variations in input
use arising from frequent price changes, or from supply restrictions.
Mehra~s work for India (8) supports the view that the new technologies have
contributed to increased yield variability at the farm level, particularly in semi-
arid regions with limited irrigation. Consequently, current efforts to breed for
more stable yields are directly beneficial to farmers.
However, even if t~e new technologies have increased yield risks at the farm
level, this need not imply that they are an important source of increased in-
stability in national cert:al production. Other factors affecting aggregate supply
include changes in interyear variability in crop areas s~wn) changes in yield
6 PINSTRUJl-ANDERS£N AND HAZ~LL

correlations between farms and crops, production expansion into riskier areas;
and increases in average yields and average areas sown.
Hazell (7,10) used statistical identities to provide an exact decomposition
of the components of change in the variance of total cereal production between
IJfC- and post-Green Revolution periods in India. He found that increased yield
variances within crops and states dif(~ctly accounted for less than 10% of the
increase in the variance of India's total cereal production. More important were
increased covariances between crop yields, and particularly between the yields
of the same crop grown in different states. A part of these increased yield
covariances is attributable to more variable yields, but part is also due to
increased yield correlations between states. In the pre-Green Revolution
period, there were stronger offsetting patterns of variation of yields between
states. Today> yields have a stronger tendency to move up and down together
over large parts of India.
This patt~rn is even more pronounced for maize in the United States (7);
interstate yield correlations increased sharply between the periods 1950-1966
and 1967-1980. In fact these increased yield correlations account for about
half of the increase in the variabiiity of total cereal production in tHe United
States between these two periods.
\Vhy have the yields of the more important cereals become more highly cor-
related between states in India and the United States? Since the phenomena in
the United States is specit1c to maize, this suggests a crop-specific cause. A
possible explanation lies in the common genetic base of most of the hybrid
maize varieties grown.
Hargrove et al. (11) report how in the spring of 1970, a mutant fOfTll of
f1elmillthosporiu11l maydis (southern corn leaf blight) struck in Florida, and
spread rapidly northwards throughout the cornbelt with devastating effects on
yields. This susceptability to H. maydis was limited to hybrids carrying cyto-
plasm associated with male sterility--a feature of considerable value to plant
breeders. Unfortunately, almost every maize farmer in the United States was
growing such varieties in 1970. The total crop loss was limited to about 15%
because of unfavorable weather conditions for H. maydis. Although an extreme
example, this episode illustrates how maize yields could have become in-
creasingly correlated over time between states as varieties be~ome more geno-
typically similar with a common susceptability to the same kinds of pest,
di~ease. and weather conditions. 3 The ris~ associated with this loss in genetic
variation has been recognized by plant breeders, and recent years have seen ~.ig­
nificant attempts to 1 everse the problem.
Another possibility is the role of increased price var~ability. Fann gate maize
prices were relatively stable around trend ill the United States over the period
1950-1972; the stand.rrd deviation was SO.20 per bushel. However, after 1972,

J Chang (12) has reported similar problems with rke in Asia.

GREEN REVOLUTION 7

their variability increased considerably; the standard deviation around trend

went up to S0.33 per bushel. Sin~e price variations affect all U,S, farmers,
any adjustmenb they make to input usage in response to price rr.ovemcnts
could lead to more highly correlated yields across states.
In India, interstate yield correlations have increased for all cereals except
sorghum and minor millets. Possibly the narrower range of genotypes for rice
and wheat resulting from widespread adoption of high-yielding varieti~s has
contributed to these increased correlations. However, other causes may be
important.
Kothare (13) provides evidence that rainfall was more erratic in many parts
of India during the more recent period of analysis, The usc of fertilizers and
irrigation on cereal crops has also increased at a time when the supplies of these
inputs have become more erratic. For example, an increase in electric irrigation
pumps since the mid-1960s has coincided with increased irregularities in elec-
tricity supplies. Since power outages affect large areas of India simultaneously,
they are likely to have a simultaneous and negating effect on crop yields in
many states.
As in the United States, cereal yields in India arc also responsive to price
signals, and increasingly so since the widespread adoption of input-intensive
technologies. The standard deviation of detrended harvest prices for rice in
\Vest Bengal (the most important rice-growing state) increased from Rs. 14.9
per quintal in 1952/1953-1964/1965, to Rs. 40.3 per quintai in 1967/1968-
1977/1978. Likewise. the standard deviation of the detrended harvest price
for wheat in Uttar Pradesh (the most important wheat-growing state) increased
from Rs. 8.2 per quintal in 1955/1956-1964/1965, to Rs. 19.8 per quintal in
1967/1968-1977/1978 (7).
Other important sources of the increased variability in India's total cereal
production were increases in the cropped are3, in the year-to-year variability
of the areas sown to indivirl ual crops, and in more positive correlations be-
tween areas sown and yields. These sources of increased production variabili~y
reflect the importan:e of increases in the gross cropped area (from 89.7 million
to 101.0 million ha between 1~52/1953-1964/l965 and 1967/1968-1977/
1978) as well as increases in mean yields in enlarging total cereal production in
India. In cOJ~trast, nearly all of the 2.7% per year growth in U.S. cereal produc-
tion since 1950 can be attributed to increased yields.
Given the potential link between the new technologies and increased in-
stability, should greater priority be given to reducing instability in agricultural
research?
In developing countries continued growth in foodgrain production is of
paramount importance, and any trade-off between breeding for growth and
breeding for stability may prove costly. But a strengthening of ongoing efforts
to develop genet;'~ r~sistence or tolerance to pests may meet growth as well as
stability Quais. Also. there are often other important sources of increased
8 PINSTRUP-ANDERSEN AND HAZELL

variability in production that would not be affected by changing research

priorities, These may be amenable to, or even the consequence of, govermnent
policy. For example, policies to provide more stable farm prices and electricity
supplies could help stabilize cereal production in India.
Appropriate policy intervention can also alJeviate some of the problems
posed by increased instability. Storage schemes and international trade can
stabilize food supplies for consumers, and credit and insurance schemes and
futures markets can help fanners to cope with yield risks. Government could
also give greater priority to increasing production in the more stable areas, and
in areas where production is not strongly correlated with other regions.

DIRECT L\tPACf ON POVERTY AND NUTRITION

A number of early studies on the impact of the Green Revolution concluded

that the rural poor did not receive their fair share of the generated benefits.
It was argued that mostly large farmers adopted the new yield-incre~sing tech-
nology, leaving small farmers unaffected or actually worse off because the
Green Revolution resulted in (a) downward pressures on the prices of the com-
modities they produced, (b) upward pressures on the prices of the inputs they
purchased, (c) efforts by large farmers to either increase rents to tenants or
force the tenants off the land, and (d) attempts by large farmers to increase
land holdings by purchasing smaller farms, thus forcing small farmers into land-
lessnes~. Furthermore, it was argued that the Green Revolution resulted in
reduced nIral employment. The net result, as argued by some, was a rapid
increase in the inequality of income and asset distribution and a worsening of
rural poverty in areas affected by the Green Revolution (14-18).
While the general validity of these conclusions has always been questioned,
recent studies have produced a sizable body of evidence which proves beyond a
reasonable doubt that they are wrong. Recent evidence clearly shows that, al-
though exceptions exist, as a general rule the Green Revolution has resulted in
a very significant improvement in the material well-being of the poor.
\Vhere did the early studies go wrong? Primarily on four accounts. First, the
studies failed to distinguish between early and subsequent adoption of new
technology. Thus, studies undertaken soon after the release of the fust high-
yieldIng rice and wheat varieties found that only large farmers adopted these
varieties together with complementary inputs such as fertilizers. \Vhile it was
tnle that early adopters were primarily large farmers, the studies failed to
recognize that small fanners would follow quickly once the uncertainty was
red uced by observing success under fann conditions.
Second, the benefits to the poor as consumers of rice and wheat through
lower prices were largely overlooked. Third, little or no attention was given to
GREEN REVOLUTION

the multiplier effects of the Green Revolution and the fi:suJting impact on
incomes of rural poor. Fourth, the impact of the Green Revolution vIas fre~
quently confused with the impact of institutional arrangcrnents, agrkultural
policies, and labor-saving mechanization. Such confusion lead to irH.;orrcct
identification of the causes of rUla) poverty and thus incorrect rccomm('nda~
lions for action to red uce such poverty.
This section provides a brief summary or current evidence of the direct
impact of the Green Revolution on the poor and how institutional and poUcy
changes may alter this impact. 4 The indirect impact is dealt with in the next
section.
Five factors are of great importance in detemlining the distribution of eco-
nomic gains from technological change in agriculture. These are (a) the nature
of the technology, (b) t he structure of the agricultural sector and particularly
the land tenure system, (c) the structure of the markets for inputs (fertilizers,
labor, ..:tc.) and credit, (d) the market for agricultural products, and (c) agricul-
ttlral policy. The impact of each of these factors is analyzed elsewhere (3) and
will only be summarized here together with evidence from more recent studies.
The poor may be affected by the Green Revolution through changes in the;r
assets, incomes. and/or the prices they pay for food. The impact on a particular
household is influenced by the extent to which it depends on rice or wheat for
its income and the importance of these commodities in the household budget.
The impact will differ among poor prod1Jcers. landless iabor, and poor con-
sumers. Recent evidence of the impact on each of these three groups is briefly
discussed below.

Impact on Poor Producers

High-yielding wheat and rice varieties have been adopted widely by producers
irrespect{ve of farm size and tenurial status (2.3, 19 -2 2). Earlier conclusions
that the Green Revolution was predominately a large farmer phenomenon were
clearly incorrect. In many, if not most, regions suited for the high-yielding
varieties, low-income fanners have adopted at least to the same extent as
larger fanners, and the most recent studies suggest that net gains -per unit of
land tend to be larger on smaller fanns (see above references). However, many
regions are not suited for high-yielding rice and wheat varieties. Thus, the
Green Revolution has contributed to a considerable change in regional income
distribution in some countries, as illustrated in India. The Green Revolution
has benefitted producers who control optinlal production environments or who
have access to such environments, irrespective of farm size. In some countries,
not least In Latin America, optimal production environments are frequently

.. See Ref. 3 for a more de~aiJcd synthesis of available evidence up to 1982.

to PfNSTRUP-ANDERSEN ANfJ HAZELL

controlled by the larger and bettef-()ff fanners. In many other areas, including
those where good soil has been distributed through land reforms, they are often
controlled by low-income famlers.
The most important lesson from these findings is that the physical produc·
tion environment, e.g., soil quality, acces-s to irrigation wa~er, ctc., is much
more important in determining ado,tion patterns than farm size. Therefore,
as long as new technology is suited only Of mainly for optimal production en-
vironments fanners without access to such environments, irrespective of farm
size, will not benefit. While a great deal of research is now under way to in-
crease the productivity of less-favored areas, available technology is still limited.
The impact of technological change on poor fanners depends very much on
institutions and policies. Although small farmers in many developing countries
generally utilize available land more efficiently than those with larger farms,
pohcymakers often see large farms as more desirable. Therefore, attempts by
early adopters of the Green Revolution technology to enlarge their farms by
land purchase or tennination of rental arrangements were supported by public
policy. Such policies were fueled by the belief during early phases of the Green
Revolution that smaller farmers would not adopt the new technology. Based
on hindsight it is obvious that what was needed during those initial phases
were policy measures and institutional changes aimed at reducing the time lag
between the adoption of the new technology on large and small farms such as
the removal of input market constraints for small famlers.
Up until a few years ago, there were indications that a combination of early
adoption generation of large economic gains among large farmers, and policies
adverse to small farmers led to increasing land concentration and increasing
numbers of landless farm workers during the initial phases of the Green Revolu-
tion. However solid empirical evidence that this is a widespread phenomenon
does not exist. While changes in land tenure have occurred, the role of the
Green Revolution is not clear. Demographic pressures, regional migration, and
other factors may have been important.

Impact on Landless Labor

The Green Revolution is based on a cOulbination of varieties with high yield

potentia!, fertilizers, irrigation, and in some cases chemical pesticides and
mechanization. S One result of this combined package has been higher labor

S Although the complete set of inputs is discussed here, it should be pointed out that- the high yield

potential of modem varieties did not depend on the adoption of all inputs, e.g., mechanization. Further-
more, as discussed below, the impact of high-yietiding varieties should be viewed separately from the
impact of other inputs such as chemicai pest control and m·::lchanization. ~fany past studies have failed to
do so and much confusion exists regarding the impact of each of the components. For example, while the
yield of modern varieties generally does not increase with increasing farm size, economic benefits from
GREEN REVOLUTiON II

productivity and increased labor demand. 1n arcas with high tlOcmpfoyment

and a highly clastic labor supply. this has resulted in a considerable expansion
in employment. In rcgicns with little unemployment and an inebstic labor sup'
ply whether existing prior to the introduction of, or brought about by, the
technology, considerable wage inCft;:ISCS have occurred . IloNCVCf. in-mig,raUon
of labor from other regions and availability of labor~saving mechanical tcch'
nology has limited such wage increases.
In many regions where technological change has made a significant impact
on priJduction real wages have increased little if at .:nI1 in spite of increased
labor dcmant! The primary reason is that these same regions have experienced
high rates of population growth. and tlH? growth in labor demand may merely
have kept pace with the growth in the labor fOfce. The important question
is how the landless would have fared III the absence of the technological
change. Clearly they would have been worse ofr.
Mechanical technology has not always reduced cmplQymenL Much of the
investment in irrigation eq uipment and in some cases investments in machines
for land preparation and harvesting has resulted in larger production increases
partly due to a shorter cropping season and dee to expansion from single to
double cropping. Shoner season varieties have also contributed to this. The
result has been increased employment. However. then;.- are many cases where
mechanization has been introduced for the explicit purpose of reducing labor
demands \vhilc the output effects have been minullal. For example, Binswanger
(23) concluded that existing studies of the effects of tractors in South Asia
failed to provide evidence that they substantially contributed to increases in
cropping intensity. yields, and timeliness while they did substitute for labor.
Investment in labor-saving technology has frequently been a reaction to labor
shortages at the local level and difficulties in labor management. Increasing
labor mobi.~ity. whether within or among countries, w'ould make such invest-
ments less attractive while increasing employment.

Impact on Poor COnSUIl1ers

Empirical evidence of consumer gains from technological change in developing

country agriculture is plentiful (24-29). The consumer gains come about be-
cause food prices are lower than they \JVould have been in the absence of the
production increases induced by technological change. Import substitution,
exports, and domestic price policies can dalnpen the price reduction. In fact.
price and foreign trade policies have been used extensively to strike a desired

certain types of mechanization may, Furthermore, benefits from modem varieties may be achieved irre-
spective of whether mechanization takes place. Yet, the negative impact of mechanization on labor
demand is often credited to modern varieties.
12

b:~hm '.1 A\l '('J:'; q~;, I' ,~','~nn!!

effects of price flUs. ,~vx hHf:H.'rs "Hid flliU:'~; HH)d
pr.:~:HJ',;~:{~h .' ; ',' tenli:~jdaJ effects on cor J'ifl' ':f':i. ~inct:' rhe Gn..>t:th Rcv{~~hl'
110.;[1 ;-.tlil",·',l(( ~ ';';~'l eC(r,:1mic :"urplus by more .:;-ffh:.)t:iut utilization of n:soUt'fCS,
and rcdu~ed unit (Ilsls. consumer l!,ains need w.H :mply prodw.:cf' loss~'s, Both
11 '\Y gain.
The distribution of C•.'0$~~~. 1ctween the poor and the rest depends
primarily 00 the am,',:;:·, ,,;em .. t group on the commodities in qUC5l!O~L
e.g., rice and wheal" Tlh:lari'~·.; th~ amount, the greater would be the gain. In
Illost cases the poo: sper.d h~$S than the bcUcr~off population on wheat and
rice ~ thuslargcf absolute g~'ins are obtained by ttw latter. However, in many
countries, the share of total incomes spent. on \vhemt and rice is larger for the
poor than th~ bctter~oJJ. Thus, real incnme gains relative to current incomes
arc larger for the PO{lr.
A su\.:!y of the population of CaH, Colombia (30) iIIustra~es these condu·
sions. In this study it was found thal lo\v·income consumers obtained larger
absolute gains than did highcr-incomte consumers from output-expahding t~ch­
nology for basic staples (cassa~r~, maize, and plantain). For all other foods
including rice and wheat, the absolute gain obtained by low-income consumers
was srnallcr than that obtained by those with higher ·ocomes. However, the
distribution of benefits from outpUit-expanding technology was less skewed
than existing income distribution for all foods. This means that in the absence
of government intervention and \vith no changes in imports or exports, new
technology would provide larger rel~l1tive benefits to the poor and improve the
existing income distribution among consumers.

[mpact on Nutritional Status

Technological change In agriculture intluences human nutrition through i [s

impact on:

I. Incomes acquired by households at risk of having malnourished or under-

nournished members
The prices they have to pay for food commodities
3. The nature of the production systems among semisubsistence fanners
4. Risk and fluctuations in food product!0n, storag(~, prices, and inc "mes
5. The nutrient composition of th(~ foods available to malnourished house-
holds
6. Household income composition, intrahousehold income and budget con-
trol, and women's time ailocation
7. Labor demand and energy expenditures
8. Infcct.ious diseases
GREEN Rf!VOLUTJON

Although the reasons for existing caloric-protdn th::fkiende') differ aJ~ilOm~

countries and population groups. low household inconu~s. insunk~(>nt food
availability. nnd high food prices arc ~ikely to be tiH~ prirnnry on(;;}< ('hangt:s,in
either of these three factors arc likely to inOuencc food COjlSlUnpliollJ, Ch~utAe$
in food supplies ~lrf."~ct the nutritionall status only to the extent that the fond
consumption of malnourished or at-risk ;ndividuals is aff("ctc<.L The dcgn.;c (0
which expanded food production is translated into expanded food consmnp~
lion by the malnourished varies greatly depending on the crop or livestock
species for which production is expanded, the nahlfie of the technology \vhkh
brought about the expansion. and who produces the increase. Thus, using total
production expansion as a proxy for nutrHional impact is likely to be mislead-
ing. \Vhat is important from a nutritional point of vic,v is not how much morc
is produced but how food consumption by the maJnourishcdis affected.
Although studies of the nutritional impact of the Green Revolution arc very
scarce. it is well known that the poor spend a large share of additional incomes
on food. Furthermore, since \vhcat and rice arc nreferred staples among the
poor in most developing countries. the poor tend to mak~ rdatively large ad-
justments in wheat and rice consumption in response to price changes. Thus,
increased incomes and reduced wheat and rice prices are likely to cause sig-
nificant increases in thcir food consumption. The extent to which these in-
creases are reHeetcd in nutritional status depends. on intraholisehold food
distribution, health, and related factors which In tum may be influenced by
technological change through changes in the intrahousehold income fonlroL
demand for \V'omcn':s; time. use of child labor. etc. 6
Since nutritional prob:ems are found among urban as well as rural poor, the
most desirable technology from a nutritional point of view is that developed
for food commodities primaril;l prod~.:ced by poor farmers or if produced by
t

better-off farmers, causes a large increase in labor denlilnd, and which occupy
a large share of the budget of poor consumers. In many developing countries~
particularly in Asia. wheat andlor rice are such commodities. In other
countries, millet, sorghum, maize. cassava. or beans would better meet these
criteria
Because of the higher producti\'ity of rice and wheat relative to other crops
for which no Green Revolution has yet occurred, many farolers have substi-
ruted wheat or ric(' for other crops on their land. Some concenl h~ls been
expressed that such substitution may have resulted in negative nutrition effects.
Substitution of wheat for pulses in India is a case in point. HO\lICver. the net
impact of this substitution has been an increase in the production of calories,

6 These relationships are further discussed In Rc'L 31.

protein, and essential amino ~ldd$ per unit of I~Uld tJ,1), T impa,:t on ahe
poor :and malnourished is not kno\1/n.
Shifts from multiple to mono(roppjn~! pr~i('nl a ri'it~ tri'it' fftW nHOJ»
effects among $cn1lsuhsistcntc r~jfnl(:r$. bmpirkal c\'ldcncc rm this loph.: 9$
Ihnitcd and more research is m:cd!l'd to a~scs'S lih: nutrition cffcds of sudl
5hirts ~uld how negative effects may be avoided "mu positive ones cnim
"t::

In particular. it \"ouM he useLai 10 furth~r slmly these and tHher on~fann

link~lgcs between Jgrjcultuml prodm:«Ion actBvHIC:-:, Ilmu lhe nutritional status 0"
(he semisubsisiCUC(; farm f~Hnjiy ~\S ,.rt of ongoing and planned farming sys¥
h.-BUS rcse~lrch,

l'UE INDUU:CT IN(OME AND E~n)LOYMENT

EFFECTS ~ IF '('rtF GHEEN nEVOl.UTION

The lnthn"d d'fCCls ~ahing lrom tc('1moloFical dtangc and agrlcult ~Hal grov,t( Ii
GH1 besubst~mu;iL ~knor Jud Ldll; {JJ,J,I» hav'c arguftd tilt" importance of
ifH:r",~ascd food prodm:tion in relaxing thl" 'V~ugl" goods' ~onstraint on economic
g,n:'HJdh In an empirkal study of agJlouHuml and industrial performance in
IndiJ. Rangar~llan (35) rOHnt~ that a I additiOn to the agricuhural grov/(h
rate 'St!rrrmiall~d a 0 5; addition to lhe gnHvth rate of industrial output, and a
CL 7 addition to the tf.!Jo,vth riih.' of miltonal incoune, At a lonal i,~v('1 Gjbb
C,6, found {hatcach 1'" increase iin agrbcuhund in(ome Hi the m:vaEcija
PnlVBKt" of Ct'rHr~ll Luzon in the Phiii!}pincs gelrH'Tated a l-~~ Increase m
employmtt>nl H] most ~ectors of the lotal nonfaon econolny. Sirnilarly, in ;3
study of te(hnoY{jgic~ii chang(' iril fin: in lhe Muda fit."glon of Malaysia. Bell d al
(31) fi)lmd ahat for each dollar of IHCC1i[]ie cn:atcd directly in agriculture by the
pn)ji:cc an additional HO (>ents or '.alm." ;;uJdiL'O Vlas created indjrcl'Uy in tiH'
!n(~li ucmfarm ~~(:nnomy
An import~HH ~lsp('d of gro\\'th ImkJgcs to the nonfann economy is that
t hey are predominantly due to increases un household consumption cxpendi~
lure. Bdl et aL report that about two-thirds of the 80 cent income multiplier
in ,\tuda was d!ue to increased rural household demands for COnStHhCr goods
and scr,."ices~ only one-third was due to agrkultures increased demands fOT
inputs and processing, transport, ;and marketing services. Gibb also found
strong employment links to the nonfood consunlcr-oricntcd sc~tors in his study
of Nueva Ecija. These finding.."i strongly support L\fcnor~s contention (38) that
be'..iuse much of the accepted wisdom on development strategy ignores these
consumption linkages. it has tenact, to seriously underestimate the potential
GREEN REVOLUTION

importance of agcrkulturc. llirschm~Hl~ for example, in hj{" infltkiltns study of

the hnportance of linkages in promoting development (39t fucu'--rl! (,nl};' on
production linkages, and he found these to be weak for agriculturt' (omp~fn;d
to most other sedors of the economy. On this basis. he rcconwl~T5dcd t
greatest priority be given to publicinvcstmcnt in nOliagrkulturc.
1n addit ion to enhancing agriculture's contribution to national economjt:
growth, Ih,,' existence of strong consml1cr expenditure linkages betw'ccn
agricultural houschjlds and the nonfarm economy is important for two other
reasons,
FirsL the income and employment g<:'ucratcd by tlH:SC linkages is predom-
inately concentrated in rural arcas. Rurally focused growth is desirable in many
countries \vhcrl' rural areas have been severely disadvantaged in the past
through ur'C,:m-biased policies (40,. Such policies have encouraged excessive
migration from rural to urhan areas and have exacerbated problems of rural
undcH'BJptoyment.
Second. t h(" k ;nds ()f goods and st>fviecs dcmum:kd arc typicaHy prod tIeed by
smail labor-intensive ·'nlcrpriscs. Thl'}' an" f,)CUSl'd on such sectors as transpor-
tation. hotels anJ restaurants. entertainment. personal services, health, dis-
tributive trades, and hOllsing and residential construction. Increased household
demands for speci~,dly agricultural products, particularly fresh fruits and veg-
etabies, and fish and livestock products, can also provide important iIlcreases
in rural employment.
Strong household lmks to lhe rUfal nonfarm economy not only help alleviate
problems of rur:lI underemployment. but. because the major beneficiarks of
lhemcre~iSed employrnent earnings are typically the POOL they also contribute
to the reduction or fura] po.. ..:rt)' and malnutntion. Survey evidence from many
(,o~lntTies confirms lhal the small farmers and landless \llOrkers obtain substan-
tial shares of their total income from nonagricultural sources. Consequently,
the beneficianes of tht inducet employment gains generated by agricultural
growth need not be limited to poor, nonagricultural households residing in
towns. Rather they have the poterHial to touch a wide ranze of occupation
groups within the poorer segments of soenety.
The indirect benefits from agricultural growth are not restricted to the poor.
They can increase the eanlings of skilled 'A'orkers as \'len as providing lucrative
returns to capital and to managerial skills. In the Muda study, for example.
BeB et at. found that the indirect benet1ts of the project were skewed in favor
of the nonfarm householdc. id the region, many of whom were relatively well
off. They also found that even among agricultural households. the landed
households fared better than the landless. The point to be made is that al-
though the indirect effects of agricultural growth do not necessarily improve
the relative distribution of income within rural areas, t1hey can still have wide-
reaching effects in aUeviatilg absolute poverty.
16 PINSTRUP-ANl()I~RSENANn If A.lELt

Determinants of the Size of the

Indirect Benefits in Rural Regions

Given the importance of household consumption expenditures in gCfH.:'mHng

income and employment multipliers. three key considerations stand out as
determinants of the size of the indirect benefits that can be induced by te(h~
nological change and agricultural growth within rural areas.
The first consideration is the amount of extm income generated among
farmers as a result of increased agricultural output. Obviously ~ the greater the
increase in farm incomes, the grcal.cr the incremental expenditure by farm
households on consumer goods and services. Agricultural growth is not always
accompanied by incn.'3ses in farm uncomcs, paniC'ularJy when the growth is
rapid and the national demand for the output is price inelastic. Some govenl~
mcnts also hold farm prices artificiaUy low in an attempt to reduce the price of
food for urban consumers.
Technological change of the Green Revolution type usually reduces the cost
of producing a unit of agricultural output. Consequently, some reduction in
prices can still be consistent with increases in farm incomes. More generally,
though. price reductions act to tral1::fer income from rural to urban areas, and
they reduce the amount of indirect growth generated within rural regions.
A second consideration is the stp~:· ufl:~ of rural household expenditure pat-
h~nlS (33), To a regional economy. ~ocal household expenditures on imported
goods «from outs"Je the region) represent a direct leakage which reduces the
size of the IOcCal iJKOH1C and employment multipliers. But as Siamwalla has
argued (41). if incremental income is spent on locally produced goods that
could be exported from the region at a constant price, that expenditure repre-
sents a loss in export proceeds and is as much a leakage as if the inoney \\teTe
.'xpendcd upon imported goods. Thus the stimulative effect of increased house-
hold expenditure on the local economy depends crucially on the expenditure
share allocated to locally produced goods and services that are not usually
exported from or imported into the region (nontradables).
For most rural regions in developing countries, lthe major output is food or
cash crops the greater part of which is exported out of the region at given
prices. In return, these areas import from outside the region manufactured
goods for investment. production, and household consumption. Nonfood
goods and ~~rvices are also produced in the towns and larger viJiages which
cater almost exclusively to local demand. Local nontradable goods m:: . also
include some specialty agricultural commodities such as fresh fruits, vegetables,
and livestock products.
Hazell and Roell analyzed nou:;ehold expenditure data from Malaysia and
Nigeria and found that the share of incremental income allocated to fural non~
tradabJes increases \vith household income and fann size (42). This suggests
GREEN RP'"·OLUTION

that within a region agricultural growth will have a grcater stimulative irnpact
on the local economy if the growth is concentrated on the larger farms. It
should be stressed though that their data did not include any really large
farmers or any really rich households. It is quite possible that the marginal
budget share for local nontradables reaches a peak for some intennediatc farm
size, and that really large farmcrs havc less desirable expenditure patterns for
r~gional economic growth. L'uger farmers may also generate other leakages
fro III the local economy. For example, they may allocate larger shares of their
income to savings which are not invested locally in goods with a high content
of local nontradables. They may also us(~ farm technologies that require larger
shares of regional imports than the technologic" l1sed by smaller farmers.
\Vhile household expenditures on local nontradablcs arc an important de-
temlinant of the size of the regional income and employment multipliers, it is
also crucial that the supplies of these goods be elastic. This leads to the third
key consideration of our argument-the suppty structure of rural regions.
If the supply structure of nontradables is inelastic, th,~n increased household
demands for these goods and services will simply increast~ prices rather than real
incomes and employment. As it happens, most nontradables are services (trans-
portation, hotels and restaurants! entertainment, personal services, health,
education~ housing. and distributive trades, etc.). These activities tend to be
labor intensive, so we should expect their supply to be morc responsive to
demand in labor surplus regions. This is nlore likely to be the case in South
Asia than in Africa. Other factors that nlay be important are the status of the
local infrastructure (roads, electricity, market centers, banking, etc.) and gov-
ernment policy towards small businesses. Again these factors tend to be more
favorable in Asia than in Africa. Since the Green Revolution has tended to be
focused in labor surplus areas, and particularly in Asia, it probably has gen-
erated substantial rounds of indirect benefits within these areas.

THE ROLE OF \VO~fEN IN TECHNOLOGICAL CHANGE

Although earlier work on the Green Revolution paid little attention to the role
of women in technological change and inlplications of their role for technology
adoption and utilization and distribution of benefits, a number of studies have
now been completed on the subject (43-47).
There are many reasons why the role of women should be explicitly ana-
lyzed. First, women playa major role in providing the additional labor required
to obtain the benefits from technological change. Failure to consider gender-
related differences in labor availability and demand may result in low adoption
rates and a lower yield impact than expected (4 9 >. In cases where additional
18 PINSTIHJP-ANnERSEN AND HAZELL

labor is needed in activities traditionally done by women it is til!! female labor

availability rather than total labor availability that is importane
A large number of studies of women's tirne allocation sho\v that women in
poor rural households are not generally underemployed. On the contrary. their
working day is usually considerably longer than that of the men. I fov:C' vcr,
much of their time is spent on low productivity but essential activities. Thus.
introduction of new agricultural technologies which require additional female
labor whether in production or processing should be associated with pro-
ductivity-increasing technology in other activities traditionally done by women.
Unfortunately, however, efforts to increase labor productivity in these ac-
tivities arc not widespread. Furthermore, attention should be paid to the im-
pact of increasing demand for female labor on the production of subsistence
commodities, child care, and other aspects of household welfare including,
of course, the burden placed on the women themselves.
The Green Revolution has dearly increased employment of the landless
poor, many of whom are women. However, labor-displacing technology has
played a greater role than productivity-increasing technology in some activities
characteristically done by women such as the substitution of chemical weed
control for hand-weeding. Where traditional women's jobs hav~ benefitted from
productivity-increasing technology they are often taken over by men.
Second, a large proportion of small farms in developing countries are headed
by women either all year or during periods when the men are away to work in
other regions. Therefore, decisions determining the adoption and impact of
technology are often made by women, a fact that should not be overlooked by
agricultural extension and other efforts to spread new technology.
Third, there is increasing evidence that household-spending behavior and the
resulting impact on welfare of individuals within the household, e.g. nutri-
tional status of children, is influenced by the distribution of income control
within the household. Thus, the impact of new technology may depend on how
intrahousehold income control is affected. Technological change resulting in
expanded production for sale at the expense of subsistence food production is
a case in point. Although total household incomes may increase, household
food consumption may fall if women are traditionally responsible for meeting
food needs through own production while the cash from sales is controlled by
the man and used for other things without sufficient increases in food pur-
chases to offset the reduction in own production.
The principal message coming from recent studies of the role of women in
technological change in agriculture is that gender specificity in labor activities,
decision-making in production and consumption, and the general level of well-
being among the rural poor is an important issue to be explicitly considered in
technology design and diffusion as well as related policies and institutional
GREEN REVOLUTION 19

changes. Ignoring these gender specificities may lead to inappropria tc technoJ~

ogies and policies from the point of view of both growth and equity.

ENVIRONMENTAL EFFECTS OF THE

GREEN REVOLUTION

\Vhile the contemporary distribution of benefits and costs among population

groups is important, the effects on intcrgenerational distribution should not be
overlooked. \Vhat has been the impact of the Green Revolution on the resource
base needed by future generations to meet their food and other needs?
A continuation of current trends of agricultural expansion into marginal
lands, rapid rates of deforestation, and overgrazing in dry areas are likely to
have severe adverse environmental consequences. Land and w1ter erosiJn and
loss of organic matter lead to land degradation and desertification, which in
tum will make it more difficult for future generations to fulfil their needs for
food, fuel wood, and other agricultural and forestry products. Existing poverty
and unsatisfied food needs, together with opportunities for quick political and
economic gains without having to bear associated environmental costs, natural-
ly lead to exploitation of the land base.
Such exploitation and the resulting degradation of the future land base may
be avoided through a combination of technological change and public policy
without serious adverse effects on short-run food supplies. New technology
facilitates higher yields on existing agricultural lands. Thus, expanding the
development and use of yield-increasing technology reduces the pressure on
new lands to meet increasing demands for food and other agricultural products.
There is little doubt that the development and use of high-yielding crop var-
ieties increased llse of fertilizers, better production practices, and other yield-
increasing factors have been of great importance in lir.'iting land degradation in
developing countries.
This very significant positive environmental effect of modern technology is
frequently overlooked. To further restrain land degradation more-not less-
yield-increasing technology must be developed and introduced. New tech-
nology may facilitate conservation of current agricultural land, and it may
assist in avoiding adverse environmental effects of incorporai~ng new land into
production. But technological change may also promote furt~er degradation
of land and water resources, e.g., waterlogging and salinization of land and ex-
cessive usage of groundwater.
In addition to accelerating the development and use of modem technology,
successful efforts to maintain or improve the productive capacity of the land
base must include appropriate public policies and investment. In particular,
20 PINSTRtJP-ANDERSEN AND HAlf3tt

policy measures are needed to ensure that long-term y,)cial costs arc reflected in
both public and private decision-making. Such polky measures arc frequently
absent in developing and developed countries.
Application of large quantities of pesticides may affect the ecological system
adversely in a variety of ways. However, a continuation of current efforts to
indude genetic pest resistance in plants, together with greater emphasis on
biological pest control and proper pesticide handling and application may
greatly reduce the environmental risks associated with pest control.
As mentioned earlier, reduction of the genetic diversity in plants is another
important environmcntal risk associated with tcchnological change. As the
diversity decreases on farms, effective steps must be taken to ensure that the
genetic material is maintained elsewhere. A considerable amount of work-
although probably not enough~-is under way in this area (12).
Environmental risks a:>sociated with the use of chemical fertilizers appear to
be rather insignificant, although excessive application rates and poor cultural
practices may result in some eutrophication of streams and lakes. However,
compared to the effects of urban and industrial sewage and wastes, fertilizer
used in agricult ure is a relatively insignificant pollutant.

LE~SONS FOR THE FUTURE

What are some of the lessons learned from recent studies and observations of
the C;reen Revolution'! First, it has become abundantly dear that the techno-
logical barriers to expanded food production among small and large farmers
in lh:veloping countries can be alleviated.
Another lesson leanled from the Green Revolution is that, while techno-
logical change in agriculture provides a vehicle for development that reaches
far beyond the more immediate goals of satisfying food and nutrition needs,
its full potential for achieving growth as well as equity goals will be realized
only if it is properly integrated into the overall development strategy and ac-
companied by appropriate public policy and institutional changes. The short-
tcnn impact on the poor is particularly sensitive to institutional arrangements
and public policies. \Vhere existing institutions favor very unequal asset and
income distributions, technological change has tended to anlplify the in-
equality. However, although the impact on the relative income distribution
varies among regions, in most cases the Green Revolution has contributed to
higher incomes of both poor and rich.
To further reduce ntral poverty and inequalities, policy measures and in-
stitutional changes should focus on the root causes of the problem, i.e., uneven
distribution of the ownership of productive resources, existing power struc-
tures, poor training and education, differential access to factor and product
GREEN REVOLUTION 21

markets, and lack of access to health facilities. Such measures might include
land reform, development of infrastructure and irrigation facilities, improved
marketing facilities, access to credit for the poor, expansions of health facil-
ities for the poor, and a series of other government intervention schemes aimed
at changing the socioeconomic environment and strengthening the human
resource. In addition to those interventions aimed at self-sustaining long-term
reductions in rural poverty, income transfers to the poor such as food and
credit subsidies are needed to ?lleviate poverty and malnutrition in the short
run. This does not mean that technological change should await such policies
and institutional changes. Technological change is needed to generate economic
surplus which together with appropriate policies and institutional changes will
facilitate growth and reduced poverty. However, technological change by itself
should not be expected to remove serious inequities. The interaction between
technological change and government policies is complex and additional re-
search on this matter is urgently needed to facilitate effective policy design in
pursuit of growth as well as equity goals.
Although the Illost obvious successes of recent technological change in agri-
culture reftr to wheat and rice grown under relatively good physical environ-
ments, on-going agricultural research is likely to result in significant yield and
production gains for other crops under less favorable production environ-
ment.s. However, if such gains are to materialize, investors in agricultural
research must make a long-tenn commitment and should not expect quick
results. Agricultural research must be viewed as an investment with a high but
long-tenn payoff. Adverse effects of failure to invest in research today may be
most severely felt 20 years into the future and beyond.
The initial successes in wheat and rice during the 1960s led to large expan-
sion,: ')f the investment in agricultural research. In addition to the International
Rice Research Institute (I RRI), CIMMYT, and CIAT which were key to the
initial rice and wheat successes, a number of other international agricultural
research institutes were created, internat.ional aid agencies made more funds
available for agricultural research in developing countries, and many develop-
ing countries expanded their research activities. However,unless new highly
visible successes outside rice and wheat come about within the not too distant
future, there is a real danger that the incentive to continue to support agri-
cultural research at current or increasing levels may begin to falter. In view of
the critical role of technological change in economic development and the need
for agricultural research to facilitate such technological change, including a
great deal of research needed to maintain the impact of current technology, it
would be very unfortunate and possibly disastrous for many countries if the
current momentum in agricultural research for developing countries were lost.
An extremely difficult situation is already upon us with respect to food
production in Sub-Saharan Africa. \Vhile the causes are complex, it may be
hypothesized that this situation could have been avoided if massive investments
22 PINSTRUP-ANDERSEN ANI) HAZELL

in agricultural research, including manpower training for the food crops and
production environments of Sub-Saharan Africa, had been made during the
past 20-25 years. The lesson to be learned is not that hindsight is better than
foresight but rather that such massive long-term investments in agricultural
research are long overdue.
A continuation of current efforts to find Hquick fixes" through price policies
and increased dependence on export crop production as a substitute rather
than a complement to long-term investments in research, training, and techno-
logical change will not lead to self-sustaining improvements in food production
and human nutrition in Africa. Neither will it provide the vehicle for economic
growth so badly needed in African agriculture. But technological change just
might provide such a vehicle.
As mentioned above, IRRI, CIMMYT, and CIAT were instrumental in bring-
ing about the Green Revolution. \Vhile many factors contributed to the success
of these institutions, including prior research on rice and wheat, the abilities
of the individual researchers, a very unbureaucratic research environment, and
sufficient research support facilities undoubtedly were of great importance.
As more international agricultural research institutes have been created and as
each institute has become larger, it is important that the high level of research
staff and the flexible and nonbureaucratic research environment so important
to success be maintained within each institute.
Furthermore, as the system of international agricultural research institutes
matures, It is very important that unnecessary and stifling bureaucratic struc-
tures not be permitted to enter the system. The individual institutes and the
Consultative Group on Ii1ternational Agricultural Research (CGIAR) have suf-
fered much less than most international institutions from the influence of and
conflict among narrow political interests. It is of paramount importance that
this situation be maintained in the future.
Closely related to this issue is the question of control over seed germ plasm.
As mentioned earlier, widespread adoption of improved varieties has reduced
the genetic diversity at the farm level. Thus, unless such diversity is maintained
elsewhere, the genetic base for future production and research will be nar-
rowed. In response to this concern the International Boare for Plant Genetic
Resources (IBPGR) was created in 1974 under the sponsorship of the CGIAR.
In addition, most of the international agricultural research institutes collect and
maintain large collections of germ plasm for the crops they work on. \Vhile
guided by scientific considerations, these efforts have been relatively free of
narrow political manipulations and conflicts and access to the germ plasm col-
lections have been free to all countries. During the last few years, however,
some developing countries have expressed serious concern about what they see
as their lack of influence in germ plasm collection and maintenance. At the
same time, private seed companies have pressed for a strengthening of patent
rights on improved -varieties and some developing countries argue that large
GREEN REVOLUTION 23

multinational companies are extracting excessive profits from sales in develop-

ing countries of improved seed based on germ plasm contributed free of charge
by these countries.
While it is clear that the development and multiplicatioJi of improved seed
will not be undertaken by private companies unless they are able to recuperate
costs and a reasonable profit, it is possible-although available evidence is
scarce-that existing laws and insufficient competition enable individual com-
panies to capture excessive profits at the expense of developing country farm-
ers and consumers.
To avoid exploitation it is important that existing germ plasm is available
to public and private institutions and companies in all countries and to pro-
mote competition among them. Effectiv~ competition among public and
private enterprises in breeding, selection, and seed multiplication should reduce
the opportunity for obtaining excessive profits. It is equally important that all
countries permit collection of germ plasm for one or more common collec-
tion(s). If, as argued by some (49), the cunent system does not meet those two
criteria, changes must be made. But it is important that such changes result in a
system that permits efficient and timely collection and retrieval as well as con-
servation of germ plasm and not a strongly bureaucratic and politicized institu-
tion which, while capable of avoiding discrimination against any country, will
be unable to provide efficient and timely support to agricultural research and
technological change for developing l..ountries.

REFERENCES

1. CIA.T, "Report on the Fourth IRTP Conference for Latin America," Cali, Aug. 10-14,
1981.
2. R. W. Herdt and C. Capule, "Adoption, Spread, and Production Impact of Modem Rice
Varieties in Asia," IRRI, Manila, 1983.
3. P. Pinstrup-Andersen, UAgricultural Research and Technology in Economic Develop-
ment," Longman, London, 1982.
4. P. Pinstrup-Andersen, "Export Crop Production," The Institute of Nutrition of the
University of North Carolina, 1982.
5. C. James, "Wheat and Maize: CIMMYT's Experience," The Courier, No. 82, Nov.-
Dec., 1983, p. 63.
6. CIMMYT Review, "CIMMYT," Mexico City, 1978.
7. P. B. R. Hazell, Amer. J. AgricuL Econom., 66(3),302 (1984).
8. S. Mehra, "Instability in Indian Agriculture in the Context of the New Technology,n
Research Report 25, International Food Policy Research Institute, Washington, D.C.,
1981.
9. R. Barker, E. C. Gabler, and D. Winkeimann, in "Food Security for Developing Coun·
tries," A. Valdes, Ed., Westview Press, Boulder, 1981.
10. P. B. R. Hazell, "Instability in Indian Foodgrain Production," Research Report 30,
Interna~ional Food Policy Research Institute, Washington, D.C., 1982.
24 PINSTRUP~ANDERSENAND Hf\ZEtL

IL T. R. Hargrove, W. R. Coffman, and V. L. Cabanitla, "Genetic f nterrclationships of

Improved Ricc Varieties in Asia," IRRI Research Paper Series No" 23. The Internation·
al Rice Research Institute, Manila, lan. 1979.
12. T. T. Chang, Science, 224, 251 ( (984).
13. N. R. Kothare, "Rainfall in India," Research Bank Staff Occasional Papers, r~escrve
Bank of India, Bom bay. Dec. 1977.
14. K. Griffin, "The Green Revolution: An Economic Analysis," UNRISD, Gencvil, 1912.
15. K. Griffin, "The Political Economy of Agrarian Ch<Jnge," Macmillan, London, 1979.
16. F. R. Fraenkel, "India's Green Revolution: Economic Gains and Political Costs,"
Princeton University Press, Princeton, 1976.
17. 1. Harriss, in "Green Revolution?:' B. H. Farmer, Ed., ~lacMillan Press, London, 19:1.

18. C. Hewitt de Alcantara, "Modernizing Mexican Agriculture:' UNRISD, Geneva, 1976.

19. ~1. Prahladachar, World Dev., / /, 927 ( 1983)
20. G. Blyn, "Economic Development and Cultural Change:' July 1983, p. 705.
21. \1. Chaudhry, Pakistan Dey. Rev., X XI(3), 173 (1982).
D. Byerlee and L. Harrington, "New Wheat Varieties and the Small Farmer," Paper pre-
sented at the conference of the International Association of Agricultural I~conomists,
Djakarta, Indonesia, Aug. 24-Sept. 4, 1982.
It. Binswanger, "Tt c Economics of Tractors in South Asia: An Analytical Review,"
The Agricultural Developmtnt Coundl and ICRISAT, 1978.
24. G. M. Scobie. "Investment in International Agricultural Research: Some Economic
Dimensions." World Bank Staff Working Paper No. 361. 1979.
25. G. \1. Scobie and R. Posada, Amer. J. Agricul. [conom., 60( 1),85 (1978).
26. R. LEvenson and P. \1. Flores, in "Economic Consequences of the New Rice Tech-
nology," IRRI, Los Banos, Philippines. 1978.
J. W. Mellor, "The Impact of New Agricultural Technology on Employment and In-
come Distribution-Concepts and Policy," U.S. Agency for International Development,
Occasional Paper No.2, Washington. D.C.. ~lay 1975.
28. \1. Akino and Y. Hayami, Amcr. J. Agricul. Econom., 5 7( 1), I ( 1975).
29. P. Pinstmp-Andersen, Eue. Rev. AgritCul. Econom., 6 (I), 17 ( (979).
30. P. P~nstrup-Andersen. "Decision-\taking on Food and Agricultural Research Policy:
The Distribution of Benefits from New Agricultural Technology Among Consumer
I Beome Strata:' Agricultural Admmistration No.4, 1977, p. 13.
31. P. Pinstrup-:\ndersen, "Incorporating Nutntional Goals into the Design of Interna~
tional Agriculmral Research," in "International Agricultural Research and Human
Nutrition," P. Pinstrup-Andersen, A. Berg. and M. Forman, Eds., IFPRI, Washington,
D.C., 1984.
32. J. G. Ryan, Ind. J. Agricul. Econom., 320),78 (1977).
33. 1. W. .\lellor and U. Lele,lnd. J. Agricul. Econom., 28(1),35 (1973),
34. U. Lete and 1. W. Mellor, Oxford Econom. Pap., 33(3), 426 (1981),
35. C. Rangarajan, "Agricultural Growth and Industrial Performance in India," Research
Report No. 33, International Food Policy Research Institute, Washington. D.C., 1982.
36. A. Gibb, Jr., "Agricultural Modernization, Non-Farm Employment and Low-Level
Urbanization: A Case Study of a Central Luzon Sub-Region," Ph.D. thesis, University
of Michigan, 1974,
37. C. BeU, P. Hazell, and R. Slade, "Project Evaluation in Regional Perspective," The
lohn Hopkins University Pres.s, Baltimore, 1982.
38. J. W. Mellor, "The New Economics of Growth," Cornell University Press, Ithaca, 1976.
39. A. O. Hirschman, HThe Strategy of Economic Development," Yale University Press,
New Haven, 1959.
GREEN REVOLUTION 25

40. M. Lipton, "Why Poor People Stay Poor: Urban Bias in World Dtv('h1fH1H.'BC" Har'iard
University Press, Cam bridge, 1977.
41. A. Siamwalla, "Growth Linkages: A Tradc~Theoretk Approach," Inlcrnaliomd Food
Policy Research I nstitutc, Washington, D.C., 1982 (Mimeographed).
42. P. B. R. Hazell and A. Roell, "Rural Growth Linkages: Household ExpendHme "a~·
tems in Malaysia and Nigeria," Research Report No. 41, I ntcmational Food Polky Hl'~
search Institute, Washington, D.C., 1983.
43. L Tinker, "New Technologies for Food Chain Activj(ies:The Imperative of Equity for
Women," Office of Women in Development, AID, Washington, D.C., 1979.
44. J. L. Unnevchr and M. L. Standford, "T(~chnology and UH~ Demand for the Women's
Labor and Management Skills in Asian Rice Farming," Paper presented at Conference
on Wo;nen in Rice FanningSystems,lRRl, Los Banos, Sept. 26-30, 1983.
45. B. Agarwal, "Rural Women and the High Yielding Variety Rice Technology in India:'
Paper presented at conference on \'."omen in RIce Farming Systems, IRRI, Philippines,
Sept. 26-30, 1983.
46. B. White, "Women and the Modernization of Rice Agriculture: Some General Issues
and a Javanese Case Study," Paper presemed at Conferem:e on Womcn in Rkc rearming
Systems, IRR1, Philippines, Sept. 26-30. 1983.
47. B. CLewis, "In"lsible Farmers: Wom(:n and the Crisis. in Agriculture;' Office of
Women in Development. AID. Washington, D.C., Apr. 1981.
48. C W. Jones, "The Impact of The SEMRY I Irrigated Ricc Production Project on the
Organization of Production and Consumption at the Imra-HoHsehoJd Level," Report
to AID, Sept. 10. 1983.
49. P. R. Mooney. "Development Dialogue:' ~\o. 1-2, 1983, p. L