Principal Paper session

Technology Adoption in Intensive Post-Green Revolution Systems
Douglas Gollin, Michael Morris, Derek Byerlee
The story of the Green Revolution is fundamentally a story about technology adoption.
Beginning in the late 1960s, modern varieties (MVs) of rice and wheat were introduced
in a number of large developing countries that were struggling to overcome food deficits,
including India, Pakistan, Bangladesh, China, Mexico, and Turkey. When grown with
reliable moisture supplies and adequate fertilizer, these MVs led to dramatic increases in
cereal production and did much to alleviate chronic food shortages.
Most of the literature on the Green Revolution and its impacts has focused on the
two decades following the initial introduction of the new technologies—roughly
speaking, the period 1960-80.1 But what has happened more recently in the intensive
systems where the Green Revolution technologies made their initial large impacts? Has
production stabilized around some dominant set of core technologies, or have intensive
systems continued to experience rapid technical change? If the latter, what has been the
Douglas Gollin is Associate Professor of Economics, Williams College and Visiting
Fellow, Economic Growth Center, Yale University ([email protected]). Michael
Morris is Senior Economist, World Bank ([email protected]). Derek Byerlee is
Adviser, World Bank ([email protected]). We gratefully acknowledge the
comments of Melinda Smale and B. Wade Brorsen. We also thank Bob Evenson for
making available his data on the diffusion of modern varieties by crop and by country.
2.
source of the technical change, and how are new technologies being generated and
diffused?
In this paper we examine technology adoption trends in intensive post-Green
Revolution systems, and we discuss how technical change continues to play an important
role in sustaining productivity growth in these systems. Focusing mainly on the irrigated
cereals-based systems of Asia where the Green Revolution has had its greatest impacts,
we suggest that a number of major changes have occurred:
•
Slow but steady gains continue to be achieved in maximum experimental yields per
unit area. These gains are still more significant when expressed in terms of yields per
day, because of shortened plant growth cycles.
•
Average yields achieved by farmers have risen significantly, including in many crops
and regions that did not benefit substantially from the early Green Revolution. More
recently farm-level yield increases in rice, wheat, and maize seem to have slowed,
although growth in total factor productivity may have accelerated.2
•
New MVs continue to replace old MVs. Successive generations of MVs have added
valuable producer and consumer traits that have served to broaden the adaptation of
high yielding varieties, stabilize production, and boost profitability:
•
Partly in response to the development of new MVs, the area that is intensively
cropped has expanded significantly. Other factors leading to the expansion of
intensively cropped area include investments in irrigation, rapid mechanization, and
changes in market access and demand.
3.
•
Emerging resource scarcities, environmental degradation, and resulting cost pressures
have led to the development and adoption of many new resource- and inputconserving management practices. These new practices complement improved
germplasm, but they involve new models of technology generation and dissemination.
•
Responding to rising wages, farmers have steadily substituted mechanical power for
labor. Mechanization in many intensive systems began with land preparation and
threshing, and now has spread to planting, weeding, and harvesting. Mechanization
has facilitated adoption of resource-conserving practices such as conservation tillage.
Taken together, these changes suggest that technological changes continue to drive
agricultural intensification.
Intensification of agriculture: An overview
Throughout history, agricultural intensification has been driven by demographic and
economic pressures. These have increased during the post-Green Revolution period.
Underlying drivers of change
From 1980 to 2000, world population grew from 4.4 billion to 6.1 billion, with 91 percent
of the increase occurring in the developing world. Rapid urbanization accompanied
population growth, resulting in a shift in food consumption patterns toward “superior”
cereals and “convenience” foods.
4.
The period from 1960 to 2000 also saw rapid growth in incomes around the world,
including in developing countries. Real per capita income in 65 developing countries for
which data are available grew at a rate of 2.8 percent annually between 1960 and 2000,
leading to a tripling of income levels (Heston, Summers, and Aten).3 These income
increases were of course, not uniform; incomes in some countries grew much faster,
while others experienced actual declines in average incomes.
Population growth, urbanization, and income gains occurred in the face of relatively
little change in the amount of land available for agriculture. The agricultural area in
developing countries grew from 2.63 billion ha in 1961 to 2.85 billion ha in 1981 to 3.17
billion ha in 2001. Between 1980 and 2000, the population of developing countries
increased by 39 percent, and income per capita rose 45 percent, but the agricultural area
grew by only 11 percent (Food and Agriculture Organization of the United Nations
(FAO)).
Strong increases in population and income, coupled with a relatively closed land
frontier, created pressures for agricultural intensification. The only alternatives would
have been decreased agricultural consumption per capita and/or increased imports of
agricultural goods. For the developing world as a whole, neither of these alternatives
occurred. On the contrary, consumption rose, and imports fell as a share of consumption.
Changing patterns of consumption brought additional intensification pressures.
During the first two decades of the Green Revolution, from 1961 to 1980, consumption in
the developing world rose from 1,927 to 2,285 calories per person per day. More than 80
percent of this increase came from growth in the consumption of crop products, primarily
cereals. During the post-Green Revolution period, from 1980-2000, consumption rose
5.
further to 2,656 calories per person per day. Only 60 percent of the increase came from
higher consumption of crop products, and relatively little of this increase (11 percent)
came from additional direct consumption of cereals. During the same period,
consumption of animal products increased by 75 percent, measured in calorie terms. This
change was driven largely by rising incomes and the high income elasticity of demand for
animal products. The growth in demand for animal products created a large increase in
the derived demand for feed grain. FAO data show that the use of grain for animal feed
nearly doubled in developing countries, from 114 million mt in 1980 (16.1% of total
grain supply) to 225 million mt in 2000 (20.1% of total grain supply). Indirect
consumption of grain thus increased dramatically, leading to further pressures for
intensification of cereal-based cropping systems.
Enabling factors
By themselves, demand-side pressures for agricultural intensification would not
necessarily have led to adoption of new technologies. Another key factor was the supply
of new technologies that allowed for growth in production, productivity, and
competitiveness of major crops.
One important source of new technologies was continuing investment in agricultural
research and development. Global investments in agricultural research totaled more than
$30 billion in 1995, of which roughly one-third consisted of public research targeted
towards developing countries (Pardey and Beintema). The absolute levels of expenditures
in developing countries grew significantly from the early Green Revolution period to the
post-Green Revolution period. By the mid-1990s, a significant quantity of private sector
6.
R&D also focused on the needs of developing countries, and this amount has by all
estimates risen in recent years.
Investments in research have contributed to technological change in intensive postGreen Revolution systems. Increasingly, however, other sources of technology have
emerged, particularly in the case of crop management technologies such as conservation
tillage and integrated pest management. Many of these knowledge-intensive technologies
that reduce input use are being disseminated through newly developed innovation
networks made up not only of traditional public research and extension organizations but
also farmers, agribusiness firms, and nongovernmental organizations.
While traditional “supply push” factors have affected the range of available
technologies, “demand pull” factors have influenced which new technologies have been
used. Price policies and market conditions have shaped patterns of technology adoption.
In addition, rising rural wages have encouraged rapid mechanization of key operations,
beginning with land preparation. In some parts of Asia, such as the Indian Punjab, nearly
all wheat and rice production is now fully mechanized. Interestingly, mechanization has
been spread almost entirely through the private sector, with little public R&D.
Adoption of improved technologies in intensive systems: Empirical evidence
In the major cereals-based cropping systems that account for most of the caloric intake in
developing countries (rice, wheat, maize), intensification has followed a reasonably
predictable four-stage evolutionary progression (Byerlee):
7.
(1) Pre-Green Revolution Phase during which traditional varieties are cultivated using
negligible amounts of external inputs. Gains in productivity are modest, and the main
source of increases in production is more extensive use of land and water resources (e.g.,
expansion in area planted, shift to more fertile land, investment in irrigation).
(2) Green Revolution Phase during which a technological breakthrough in the form of
input-responsive MVs provides the potential for a dramatic increase in land productivity
through substitution of external inputs, especially fertilizer, for land.
(3) First Post-Green Revolution Phase (Input Intensification Phase) following the initial
adoption of MVs, during which farmers improve allocative efficiency by increasing the
use of purchased inputs (e.g., fertilizer) and capital (e.g., tube wells, machinery).
(4) Second Post-Green Revolution Phase (Input Efficiency Phase) beginning after input
use has reached high levels, during which farmers use improved information and
management skills to substitute for higher input use, leading to higher technical
efficiency in the utilization of inputs while contributing to the sustainability of the
resource base.
In many intensive systems, observed changes in output growth, input use, and total factor
productivity have been consistent with this stylized view of technical change.
Adoption of modern varieties (MVs)
The adoption of agricultural technologies during and after the Green Revolution is well
documented. Evenson and Gollin (2003) summarize surveys of MV adoption by region
for 11 important food crops. These surveys focus specifically on the adoption of MVs
8.
produced by public research organizations, particularly the international agricultural
research centers. Data on adoption of MVs are more readily available than data on the
adoption of other improved technologies, reflecting the fact that MV adoption is more
easily observed and measured.
The popular narrative of the Green Revolution accurately describes the initial spread
of MVs into favorable production environments. Starting in the early 1960s, MVs of rice
and wheat were adopted by farmers in South and Southeast Asia and Latin America.
Adoption was rapid, though limited in terms of area. Yield gains were impressive; under
the most favorable conditions, yields of rice and wheat doubled or tripled.
Less appreciated, however, is the fact that the diffusion of MVs has continued
steadily for more than 40 years. In fact, the post-Green Revolution period has witnessed
greater MV diffusion than the earlier period, for almost all crops and regions (Evenson).
Even in some crops and areas that first received the Green Revolution, the later period
witnessed greater diffusion (table 2). For example, Evenson estimates that 36 percent of
South Asia’s rice area was planted with MVs by 1980; that figure doubled to 71 percent
by 2000. For East and Southeast Asia, 41 percent of the area was devoted to MVs in
1980; by 2000, the figure had risen to 81 percent.4 Much of the expansion in the use of
MVs in this period was in less favorable areas in terms of water control and moisture
supply (e.g., rice in eastern India, wheat in the Middle East and North Africa, and maize
in Latin America and Sub-Saharan Africa).
Adoption of improved germplasm continues to be an important factor driving
productivity growth in intensive systems, as evidenced by the steady replacement of old
MVs by new MVs. Often, however, the characteristics of new MVs differ from those of
9.
old MVs. High yield potential and resistance to biotic and abiotic stresses continue to be
valued, but other traits have also become popular among farmers: agronomic
performance, industrial quality, nutritional content, and perceived healthfulness. Thus,
plant breeding research has continued to add value to agriculture, though not necessarily
through increased crop yields.
Beyond germplasm: Adoption of crop management technologies
Most MVs perform well relative to traditional varieties even under unfavorable
production conditions, but they express their full yield potential only with favorable
management. Because of their responsiveness to inputs, MVs became an important
catalyst for the adoption of fertilizer and irrigation (Morris and Byerlee).
During the 1970s, 1980s, and 1990s, fertilizer was the largest source of growth in
food production in developing countries, particularly in intensive rice and wheat systems.
Use of fertilizer was further encouraged by a prolonged decline in global fertilizer prices
that began during the mid 1970s, and also by subsidies introduced in many countries.
Because fertilizer is a land-saving technology, its use was concentrated in the land-scarce
intensive systems of Asia. From a very small base in the 1960s, fertilizer use in Asia
expanded to the point that it now accounts for nearly half of total world fertilizer
consumption. In many irrigated lowland areas within Asia, fertilizer application levels are
now at or above recommended levels.5
The diffusion of MVs also stimulated increased investment in irrigation, leading to
significant growth in irrigated area during the 1960s and 1970s in Asia, especially
through the adoption of tubewells by small-scale farmers. Growth later slowed, for two
10.
main reasons. First, many irrigation systems became degraded through lack of
maintenance, so irrigation investment was diverted to rehabilitation of existing systems,
rather than to construction of new ones. Second, most of the areas that could be irrigated
at comparatively low cost had been utilized, so further expansion became technically
more challenging and correspondingly uneconomic, especially with the downward trend
in cereal prices.
The introduction and spread of MVs also precipitated a sharp increase in the use of
pesticides, especially on rice. Intensive year-round monocropping of rice led to a rapid
buildup in insect populations, as many of the early rice MVs lacked resistance to
important pests. After increasing steadily for several decades, pesticide use on rice
reversed course and started to decline as health and safety effects became more apparent
(Raheja).
For nearly three decades following the initial spread of MVs, yield gains in
intensively cropped systems were accompanied by input intensification. Eventually,
however, input intensification strategies were undermined by at least four developments.
First, in many intensive irrigated systems, rapidly diminishing returns to additional
applications of purchased inputs set in. Second, over time intensive cultivation practices
were noticeably degrading the natural resource base, with consequent negative impacts
on production.6 Third, agricultural intensification was increasing the demand for critical
production factors – especially water, land, and labor – and scarcities were beginning to
emerge in many areas. Fourth, in an effort to reduce the unsustainable fiscal drain caused
by input subsidies and producer price supports, many governments enacted agricultural
policy reforms, and these reduced the incentives to use higher levels of purchased inputs.
11.
Singly or together, these four developments altered crop management strategies in
many intensive systems. Where previously farmers had sought to increase profits by
raising yields and production, now the goal was to increase profits by reducing input use.
From the mid-1990s, farmers in many intensive systems started adopting practices
designed to increase the efficiency of input use. Conservation tillage, which had been
around for decades but which had never enjoyed widespread adoption, came back into
favor. In developing countries, interest in conservation tillage centered initially on largescale commercial systems in South America; from there it spread to northwest Mexico,
South Asia, and East Asia (Ekboir). Other input efficiency-enhancing technologies taken
up with increasing frequency included laser land leveling and bed planting (which reduce
irrigation water requirements), and to a lesser extent, integrated pest management (which
reduces the cost of pest control). Practices associated with so-called “precision
agriculture,” such as the use of leaf color to fine-tune fertilizer dosing, also started to
attract increased interest.
Although land productivity growth has been the driving force in overall productivity
growth in most Green Revolution settings, it is likely that labor productivity growth made
possible by mechanization now accounts for the largest share of overall productivity
growth (Murgai, Ali, and Byerlee). Some new crop management technologies are labor
saving (e.g., conservation tillage), but many others are more labor intensive (e.g.,
integrated pest management). A combination of labor intensity and knowledge intensity
may be slowing the adoption of these technologies at a time of rising rural labor costs.
Likely future developments: The road ahead
12.
Improved germplasm and improved crop management methods will continue to drive
productivity increases in intensive systems. In future the two types of technologies are
likely to prove complementary in the same way that early MVs were complementary with
fertilizer. Over time, MVs will remain a major source of productivity growth in intensive
systems, but improved crop management practices are likely to increase in importance
relative to earlier periods. Many intensive systems in Asia are currently experiencing a
management revolution, as evidenced by the spread of conservation tillage, laser land
leveling, bed planting, and integrated pest management. What these management
practices have in common is that they are knowledge-intensive, which means that if they
are to be adopted successfully, farmers require well-developed decision-making skills,
strong management capacity, and ready access to the latest technical information.
Producers will rely on multiple channels to supply new technologies and
information. Assuming appropriate incentives are provided, the private sector will play a
central role in agricultural innovation systems. The public sector will then be able to
focus on strategic research and public good areas, such as natural resource management
and environmentally-friendly technologies. A major question for public research systems
will be how to upgrade their capacity and effectiveness to meet these new demands.
Many emerging crop management technologies substitute farmers’ knowledge and
skills for inputs. Private firms have few incentives to carry out research on these
technologies, because they offer limited opportunities to generate commercial profits. At
the same time, most public research institutes are poorly equipped to carry out the
decentralized adaptive research needed to fine-tune crop management practices at the
local level. One approach is to have multi-disciplinary teams of researchers who work
13.
with farmers in exploring innovative management technologies (Hobbs and Morris).
Another approach is to focus on upgrading farmers’ skills through education and
information diffusion. Either approach will require continuing flows of resources.
Given the complexity of many emerging crop management practices, technology
transfer will pose a much greater challenge than in the past (when the emphasis often was
on strengthening input distribution systems). The increasing reliance of farmers in
industrialized countries on computerized planning systems and on farm- and cropmanagement software provides a glimpse of where agriculture is heading in developing
countries. In the many developing countries where landholdings are too small to justify
the use of such technologies at the individual farm level, ways will have to be found to
employ them at the village level. Farmers will increasingly turn to a range of sources for
obtaining improved technical information, including the traditional extension system,
input dealers, NGOs, farmer organizations, and even paid consultants. Remaining public
extension services will have to be considerably upgraded, and field staff will need
training and equipment to bring better decision aids to farmers.
Looking ahead to the next 10-20 years, it seems likely that the cost of agricultural
labor will continue to rise, further strengthening the demand for labor-saving
technologies. In many post-Green Revolution areas, the agricultural labor force will
begin to decline, leading to farm consolidation and increased demand for mechanization.
While the challenges facing intensive systems are large, it is striking that the natural
evolution of these systems has already alleviated many of the popular fears and
skepticism voiced concerning the original Green Revolution. The lesson that emerges—
and it is essentially a hopeful one—is that as long as adequate funding is available to
14.
support innovation, new institutions and technologies are likely to evolve in response to
emerging problems and challenges.
15.
Footnotes
1.
As used here, “Green Revolution” refers to the period 1960-80, and “post-Green
Revolution” refers to the period after 1990. Many scholars do not make this
distinction and think of the Green Revolution as encompassing a single period
running roughly from 1960 to 2000.
2.
See tables 3 and 4. For evidence on TFP growth, see Murgai, Ali, and Byerlee.
3.
The growth rate is based on aggregate per capita income in 1960 and 2000 for all
developing countries (real per capita income below $9,000) with data. These income
levels are derived by comparing levels multiplying real chain-weighted per capita
income for each country by its population, summing across countries and dividing by
total population. The growth rate is annual compound (exponential) growth rate.
4.
For wheat in Latin America and South Asia, the diffusion was far more complete by
1980; however, area of MVs in the Middle east and North Africa doubled from 19802000.
5.
Outside of Asia, fertilizer use is lower, especially in large areas in Sub-Saharan
Africa and parts of Latin America where extensive production methods prevail.
6.
Ali and Byerlee estimate that one-third of the gains in total factor productivity in the
Pakistan Punjab, in the post-Green Revolution period, were offset by resource
degradation.
16.
References
Ali, M. and D. Byerlee. 2002. “Productivity Growth and Resource Degradation in
Pakistan’s Punjab: A Decomposition Analysis.” Economic Development and
Cultural Change 54(2002): 839-64.
Byerlee, D. “Technical Change, Productivity, and Sustainability in Irrigated Cropping
Systems of South Asia: Emerging Issues in the Post-Green Revolution Era.”
Journal of International Development 4(1992): 477-496.
Ekboir, J. “Developing No-Till Packages for Small-Scale Farmers.” Part 1 of World
Wheat Overview and Outlook 2000-2001. Mexico, DF: International Center for
Wheat and Maize Improvement (CIMMYT), 2002.
Evenson, R. E., and D. Gollin, eds. Crop Variety Improvement and Its Effect on
Productivity: The Impact of International Agricultural Research. Wallingford,
UK: CABI, 2003.
Evenson, R. E. “Modern Varieties and Development.” Unpublished data set, Economic
Growth Center, Yale University, 2005.
Food and Agriculture Organization of the United Nations. FAOSTAT data,
(http://faostat.fao.org/), May 2005
Heston, A., R. Summers, and B. Aten. 2002. Penn World Table Version 6.1,
(http://pwt.econ.upenn.edu/php_site/pwt61_form.php), Center for International
Comparisons at the University of Pennsylvania (CICUP), October 2002.
Hobbs, P., and M. L. Morris. “Meeting South Asia’s Future Food Requirements from
Rice-Wheat Cropping Systems: Priority Issues Facing Researchers in the Post-
17.
Green Revolution Era.” Working paper 96-01, Natural Resources Group,
International Center for Wheat and Maize Improvement (CIMMYT), 1996.
Morris, M. L., and D. Byerlee. “Maintaining Productivity Gains in Post-Green
Revolution Asian Agriculture.” In C. K. Eicher, and J. Staatz , eds. Agricultural
Development in the Third World (3rd Edition). Baltimore: Johns Hopkins
University Press, 1998.
Murgai, R., M. Ali, and D. Byerlee. “Productivity Growth and Sustainability in PostGreen Revolution Agriculture: The Case of the Indian and Pakistani Punjabs.”
World Bank Research Observer 16(2001):199-218.
Raheja, A. K. “Integrated Pest Management in South and South-east Asia.” In A.
Mengech, K. N. Saxena, and H. N. B. Gopala, eds. Integrated Pest Management
in the Tropics: Current Status and Future Prospects. Chichester: John Wiley,
1995.
18.
Table 1. Area Planted to Modern Varieties (% of Total Area Harvested)
Sub-Saharan
Africa
East & SE
Asia and
Pacific
Latin
Middle East & America &
South Asia North Africa Caribbean
Crop
Year
Rice
1960
0.0
0.0
0.0
0.0
0.0
1965
0.0
0.3
0.0
0.0
1.5
1970
0.0
9.7
10.2
0.0
4.7
1975
0.0
27.0
26.6
0.0
9.6
1980
3.1
40.9
36.3
2.2
16.2
1985
8.6
54.1
44.2
3.3
20.4
1990
12.3
63.5
52.6
4.3
27.8
1995
19.9
71.1
59.0
7.3
30.6
2000
31.0
80.5
71.0
10.4
32.3
1960
0.0
0.0
0.0
0.0
0.0
1965
0.0
0.0
1.7
0.0
0.0
1970
0.4
0.0
39.6
7.6
11.4
1975
1.9
14.8
72.5
26.1
35.4
1980
4.1
27.5
78.2
33.8
61.3
1985
4.0
34.3
82.9
40.7
71.1
1990
6.3
58.7
87.3
43.8
79.3
1995
33.0
78.8
90.1
58.9
88.0
2000
47.4
89.1
94.5
69.1
93.2
1960
0.0
0.0
0.0
Na
0.0
1965
0.0
0.0
0.0
Na
0.0
1970
0.0
16.2
17.1
Na
1.6
1975
0.0
39.5
26.3
Na
4.7
1980
0.4
61.7
34.4
Na
11.2
1985
3.7
65.9
42.5
Na
19.4
1990
7.5
73.0
47.1
Na
27.0
1995
13.1
83.2
48.8
Na
40.1
2000
16.8
89.6
53.5
Na
56.5
Wheat
Maize
Source: Evenson (2005).
19.
Table 2. Average Yields, Major Cereals, Developing countries (kg/ha)
Crop
1961
1970
1980
1990
2000
Cereals, Total
1,115
1,482
1,874
2,426
2,724
Maize
1,128
1,494
1,969
2,447
2,766
Millet
578
737
650
769
739
1,756
2,276
2,670
3,468
3,798
Sorghum
685
859
1,058
1,084
1,136
Wheat
775
1,124
1,565
2,289
2,651
Rice, Paddy
Source: FAOSTAT.
20.
Table 3. Annual Yield Increases, Developing Countries, 1961-2000 (%)
1961-70
1970-80
1980-90
1990-2000
Cereals, Total
2.9
2.4
2.6
1.2
Maize
2.8
2.8
2.2
1.2
Millet
2.5
-1.2
1.7
-0.4
Rice, Paddy
2.6
1.6
2.6
0.9
Sorghum
2.3
2.1
0.2
0.5
Wheat
3.8
3.4
3.9
1.5
Source: FAOSTAT.