Occasional Papers Series
Economics and Sector Planning Division
Office of Agriculture
Technical Assistance Bureau
U.S. Agency for International Development
Washington, D.C. 20523
* : I
ECONOMICS AND SECTOR PLANNING DIVISION
Occasional Papers of the Economics and Sector Planning
Division of AID's Technical Assistance Bureau are circulated
for informational purposes. These papers are designed to
stimulate and serve as background for discussion on topics
of current interest to AID. They represent the views of the
authors and are not intended as statements of Agency policy.
THE IMPACT OF
EMPLOYMENT AND FOOD PRODUCTION
William C. Merrill
Occasional Paper No. 1
Economics and Sector Planning Division
Office of Agriculture
Technical Assistance Bureau
U.S. Agency for International Development
Washington, D.C. 20523
TABLE OF CONTENTS
SUMMARY OF MAJOR CONCLUSIONS AND RECOMMENDATIONS --------- 1
I. TYPES OF FARM MECHANIZATION -------------------------- 4
II. THE ROLE OF MECHANIZATION IN AGRICULTURAL DEVELOPMENT 5
III. THE EXTENT OF AGRICULTURAL MECHANIZATION ------------- 6
IV. MECHANIZATION AND YIELDS ----------------------------- 8
V. MECHANIZATION AND INTENSITY OF LAND USE -------------- 13
VI. MECHANIZATION AND ON-FARM EMPLOYMENT ----------------- 15
VII. MECHANIZATION AND OFF-FARM EMPLOYMENT ---------------- 26
VIII. APPROACHES TO INCREASING POWER AVAILABLE
TO AGRICULTURE ------------------------------------- 29
SUMMARY OF MAJOR CONCLUSIONS AND RECOMMENDATIONS
The following major conclusions are either documented in this
paper or can be documented from data in other studies of agri-
1. The horsepower per hectare available to the agricultural
sector is extremely low in most LDC's. Mechanical horsepower
per cultivate hectare usually is less than 0.3 hp in Latin
America, 0.1 hp in Asia (excluding Japan and Taiwan), and
frequently below 0.02 hp in Africa. Increases in agricultural
production will require increases in agricultural power and
better utilization of available energy sources.
2. Mechanization is the only viable means of increasing agri-
cultural horsepower per hectare over the long run. As the quality
and quantity of arable land per capital diminishes, the food energy
for men and animals will become increasingly expensive. Mechan-
ization, in a sense, reduces the demand for food.
3. Fossil fuels are likely to be the main energy source for
mobile agricultural equipment for a least the next twenty to
thirty years. Conservation of fossil fuels will require increased
use of other energy sources (solar radiation, wind, gravity,
agricultural by-products) for stationary equipment such as grain
dryers and irrigation pumps.
4. In its early stages, mechanization usually has very little,
if any, effect on crop yields. In actual practice, it appears
unlikely that deeper plowing, better weed control, or improved
grain harvesting resulting from mechanization will increase
yields by more than 10 percent.
5. In its early stages, mechanization may be associated with
a slight increase in multiple cropping and some minor adjustments
in the types of crops grown. The introduction of tube wells and
irrigation pumps appears to have a much greater effect on cropping
intensity than other types of mechanization. In general, the
observed increase in cropping intensity which can be attributed
to mechanization probably is seldom greater than 10 percent. In
most cases mechanization is "pulled in" by increases in the in-
tensity of cropping or its potential--rather than multiple
cropping being "induced" by mechanization.
6. Mechanization which replaces animal power usually results
in a reduction in labor inputs. The amount of the reduction
in labor inputs depends on the particular crops, farm size,
and extent (as well as type) of mechanization. During the early
stages of mechanization of grain production, labor requirements
may be reduced by as much as 30 to 40 percent.
7. Non-agricultural employment engaged in the manufacture,
distribution, repair and maintenance of farm equipment replaces
only a small part of the displacement in on-farm employment
resulting from mechanization. Rough estimates based on the
limited data available suggest that each man-year in farm
employment created by the agricultural machinery industry
results in a displacement of on-farm employment of at least
20 man-years over the life-time of the equipment produced.
8. Tractorization programs may result in a gradual increase
in the size of land holdings and the displacement of tenants
or farm workers unless there are institutional factors or
government policies to prevent or discourage such trends.
9. Government policies and programs to promote mechanization
through subsidized interest rates, favorable import arrange-
ments, or increased credit availability can cause a significant
increase in the rate of mechanization and are likely to benefit
large landholders more than others.
10. The mechanization of agriculture is a continuous and inevitable
process in economic development but one whose speed and direction
can be altered by public policies and programs. There is no
single indicator as to whether or not a particular machine is
appropriate for a particular nation as a starting point, however,
it is useful to ask. "Can a particular machine be produced
profitably domestically without substantial government subsidization?"
The following recommendations are based on the assumption that
developing nations are interested in increasing agricultural
output while at the same time minimizing unemployment and, to
some extent, improving the distribution of income. They also
reflect a basic assumption that increased mechanization will
be necessary in the long run if rural welfare is to be improve-
1. Government should not subsidize the manufacture, importation,
purchase or use of tractors through special exchange rates or
low-cost credit programs. Such programs have seldom benefitted
the majority of the farm population or led to substantial increases
2. Governments should support the development of agricultural
machinery which can be used profitably by a large proportion of
the nation's farmers and, to a large extent, can be produced
domestically without government protection or subsidization.
3. Highest priority should be assigned to the development of
small scale equipment which can effectively utilize non-fossil
fuel energy. (As examples: the use of wind for pumping water
or the use of rice hulls for drying grain.)
4. The development of low cost irrigation pumps and other
mechanical devices to improve the effectiveness of water control
systems should also receive high priority. Water control is a
critical element in multiple cropping systems. Furthermore, water
is likely to become an increasingly scarce resource; one not to be
wasted with inefficient irrigation systems.
5. Additional research should also be undertaken on ways to
improve the effectiveness and efficiency of small scale farm
equipment. (Can plows or rotary tillers be designed, or re-
designed, which will require less energy, smaller tractors, to
effectively cultivate land? Can small threshers, dryers, and
grain mills be improved to further reduce grain losses? etc.)
6. Agricultural mechanization should be viewed as a part of
modern agricultural production systems. Mechanization alone
has little effect on yields or cropping intensity and usually
will result in a reduction in employment per unit of output.
Combined with new biological and chemical technologies, however,
mechanization may enable more precise timing of operations and
application of chemical inputs so that the total biological,
chemical, and mechanical package results in an increase in output
per acre year with little, if any, reduction in total employment.
The mechanization of agriculture has contributed tremendously
to the economic progress of the more developed economies. Advances
in mechanical technology have enabled the introduction of improved
cultivation practices, the effective use of modern inputs, and
increases in farm size. This usually has resulted in a major
movement of people out of farming and substantial increases in
farmers' productivity. Whether the farm mechanization strategies
followed by the more developed nations are suitable for developing
nations, however, is questionable.
I. Types of Farm Mechanization
Farm mechanization is the process of introducing tools, machinery,
and equipment for producing agricultural products. Such implements
basically transform energy and focus it on a particular task.
Mechanization can result in the more effective utilization of
energy or the replacement of one energy source by another. A
change of energy sources may have significant social and economic
effects, especially when it occurs rapidly. Government policies
can affect the type and speed of mechanization as well as the
distribution of its social and economic benefits and costs.
A farmer's technical opportunities for mechanization depend partly
on whether he produces grains, livestock, fruits, vegetables, tree
crops, flowers, tobacco, or sugar crops. Government programs to
promote farm mechanization have emphasized grains. Rice, corn,
and wheat usually are the most important food crop, frequently
are produced by a large proportion of a nation's farmers, and offer
a wide range of technical alternatives for mechanization. The data
available on the impact of technology applies to the mechanization
of grain production.
Producing grains involves land preparation, water control, planting,
weed and pest control, harvesting, and some on-farm processing.
Each operation requires a different type of mechanization but the
mechanization employed in one operation is seldom independent of
that used in other operations. This makes it extremely difficult
to estimate either the output or employment effects of mechanizing
a particular operation. The introduction of tractors for land
preparation, for example, may have little effect on output unless
water availability is increased with tube wells. The investment
in tube wells, on the other hand, may yield the greatest returns
only when tractors and associated equipment can be used to reduce
the time required for land preparation and harvesting thereby
allowing multiple cropping.
The set of mechanical alternatives for producing grain contains
numerous subsets defined for particular grains and specific geo-
graphic and climatic environments. Each subset involves a
different combination of labor utilization and yields. Only a
few of these subsets have been studied carefully. The majority
of the studies have been concerned with the introduction of two-
wheel tractors (7 to 15 hp) for rice production and the use of
larger tractors (30 to 70 hp) for rice, cotton, and wheat pro-
duction. Most studies emphasize the early stages of mechaniz-
ation and are concerned particularly with how a new machine
affects early adapters' decisions, use of resources, and income
over a three to five year period. The short and intermediate term
impacts of mechanization may be quite different from the long term
impacts. Nevertheless, it is the three to five year impacts which
usually are of greatest interest to ministries of agriculture and
international lending agencies.
II. The Role of Mechanization in Agricultural Development
Farm mechanization frequently is equated with modernization and
modernization, in turn, with high productivity of the land and
labor resources employed in agriculture. In the process, farm
mechanization frequently is viewed as an essential part of, if
not the first step towards, agricultural development. It is
argued that mechanization can, and will, increase agricultural
output and employment by bringing more land into production,
increasing multiple cropping, and improving cultivation
practices. There is no way to prove or disprove such a general
claim. The available studies on specific mechanization programs,
however, show clearly that the benefits of mechanization are not
obtained without costs and that the costs at times exceed the
Over the long run the mechanization of agriculture permits
structural change in an economy by allowing labor to be shifted
to other activities without a resulting reduction in agricultural
output. This process is readily documented where mechanization
and structural changes have occurred relatively rapidly as in
the United States. It is less easily observed where population
has increased more rapidly than the rate of mechanization. In
such cases the proportion of the total labor force employed in
the agricultural sector may change very little if at all as
mechanization occurs. It is always difficult to determine whether
labor is being "pulled out" of the agricultural sector by external
forces or "pushed out" by mechanization. Machines usually are
simply "enabling devices" which allow structural changes to occur.
In other words, the machines themselves do not "induce" or "cause"
much of the structural change.
From the viewpoint of the individual production unit, the
decision to substitute machinery for labor or animals may be
a response (a) to changes in the relative prices of inputs
or (b) to a shift in the production function which increases
the marginal productivity of machinery. The availability of
new machines may allow farmers to use their existing resources
more intensely, change the timing of farm operations, introduce
new crops, increase the size of their farms, and change the
composition of labor inputs. Numerous efforts have been made
to measure the effect of mechanization, especially the intro-
duction of tractors, on yields, the intensity of land use, and
on-farm employment. Several of these studies will be discussed
Agricultural mechanization may directly effect the non-farm
population by creating new jobs in the manufacture and main-
tenance of farm machinery and through the development of new
job skills. The development of an agricultural machinery
industry also has both direct and indirect effects on other
industries. The off-farm effects of agricultural mechaniz-
ation have not been as carefully studied as the on-farm effects.
Much of the analysis has been concerned primarily with estimating
the off-farm employment associated with farm mechanization.
III. The Extent of Agricultural Mechanization
The FAO estimates of mechanical horsepower per hectare in 1967-68
for selected countries are presented in Table I. They range from
0.01 in India to 3.09 in Japan. Human power available to agri-
culture for the countries considered in Table I ranges from 0.10
to 0.25 horsepower per hectare depending primarily on population
densities. Animal power in most of these countries is in approx-
imately the same range (Stout and Downing, 1974). Agricultural
power (human, animal and machine) available in Africa has been
estimated to be as low as 0.05 horsepower per hectare; far below
that available elsewhere (Giles, 1967). The high level of
mechanical power available in Japan reflects the widespread use
of small two-wheel tractors and intensive land utilization.
TABLE I: MECHANICAL HORSEPOWER AVAILABLE IN SELECTED COUNTRIES
Country Cultivated Hectare
West Germany 1.94
United Kingdom 1.57
United States 0.96
China, Rep. of 0.15
1/ Most estimates based on 1967-68 data.
a. Latin American estimates are based primarily on FAO data
reported by K.C. Abercrombie, "Agricultural Mechanization and
Employment in Latin America," in Mechanization and Employment
in Agriculture, Geneva, ILO, 1973.
b. Data on other countries based on FAO data reported by
B.A. Stout and C.1M. Downing, Selective Employment of Labor
and Machines for Agricultural Production, Michigan State
University, Institute of International Agriculture, Monograph
No. 3, April 1974.
IV. Mechanization and Yields
During the early stages of mechanization most new machines have
little effect on yields. Although some studies report a high
correlation between horsepower per hectare and yields or find
that farms using tractors have higher yields, a closer look
usually reveals that the higher yields are the result of using
improved varieties, more fertilizer, better water control, or
improved cultivation practices rather than machines per se. It
appears that yield increases accounted for by mechanization alone
are seldom greater than 10 percent. Many researchers assign
almost all yield increases to other technological changes which
frequently occur simultaneously with, but independent of, mechan-
ization. Mclnerney and Donaldson (1973), for example, analyzing
the consequences of farm tractors in Pakistan, conclude that there
is "no convincing evidence of beneficial changes in crop yield
associated with tractor use" and that the observed yield increases
could "be attributed largely to increased fertilizer use."
One of the most careful surveys of available evidence on the
relations between tractorization and crop yields was carried
out by Singh and Chancellor (1973). The results of the main
studies which they reviewed are presented in TableII. Some of
the studies are based on field experiments; others on surveys
of farms using different power sources. At first glance the data
suggest that the crop yields on tractor farms are slightly higher
than those on bullock farms in many, but not all, instances.
Statistical analysis of the sample data underlying TableII reveals
that the yield differences are rarely significant at the 10 percent
level of significance.
The field experiment data from the Philippines in Table III show
a 12 percent reduction in sweet potato yields when two-wheel
tractors are used to replace carabao. The use of hand tractors
had little effect on rice yields but substantially reduced yields
of cowpeas interplanted with corn while having only a slight negative
effect on corn yields.
COMPARISON OF YIELDS UNDER BULLOCK AND TRACTOR USE
(1949 50) 1/
(1952 54) 2/
West Pakistan (1969)
Philippines (1967) 4/
Taiwan (1970) 5/
Department of Agriculture, Federation of Malaya, 1953.
Ahmad, B., 1970.
IRRI, Annual Report, 1967.
Hu, C.H. Chuen-Bang Young, and Ten Kian Huang, 1971.
Complete citations of references are presented in the bibliography.
TABLE III: COMPARISON OF YIELDS FROM THREE POWER SOURCES,
Hand Hand Tractor-Carabao
Crop Labor Carabao Tractor Ratio
(1) (2) (3) (3) + (2) as %
Rice (tons/ha) 2.3 2.4 2.4 100
Sweet Potato (tons/ha) 9.4 9.4 8.3 88
Cowpea with Corn (tons/ha)l.2 1.1 0.8 73
Corn with Cowpea 32 31 30 97
(1000 marketable ears/ha)
Source: Gordon R. Banta, 1973.
Many of the tractorization programs of the 1960's were promoted with
the claim (or assumption) that increased mechanization would result
in higher crop yields. In many cases the claim appears to have be-
come a belief and the belief a part of the conventional wisdom. There
are several reasons why yield increases due to mechanization may not
be observed during the early stages of mechanization. First, the
introduction of tractors, with other technological inputs unchanged,
merely represents a substitution of one power source (tractors) for
another (usually animals). Second, farmers seldom mechanize all of
their cultivation operations during the early stages of mechanization.
Land preparation usually is the first operation mechanized while other
operations continue to be performed with animal power. More complete
mechanization may be required to obtain a significant increase in
yields. Mechanization of secondary operations, for example, may
allow more precise application of fertilizers, pesticides and herbicides
and result in higher yields. At least part of such yield increases
should be attributed to mechanization. Third, the implements or
cultivation procedures used in land preparation in actual practice
may not result in a significantly better seed bed than obtained prior
to mechanization. The farmer, in an effort to reduce costs or save
time, may not fully utilize the additional power which tractors provide
or may farm soils where deeper plowing is of little value. When
mechanization does result in better seed beds, this may reduce the
need for, or cost of, later cultivation operations but have little
effect on yields. Deeper plowing, for example, may reduce weed
growth and as a result reduce weed control costs without any effect
There are a large number of economic, social and technical factors
which affect the way machines are used during the early stages of
mechanization and thus the impact they may have on yields. Farmers
seldom switch from animal to machine power with the expectation of
higher yields from the change in power sources alone. They may,
however, recognize that in some cases machines will enable them to
use chemical and biological technologies that will increase yields.
Land preparation is one of the most difficult operations, it is
hard work for both the farmer and his animals. The timing of land
preparation to allow for compatibily with irrigation schedules and
planting seasons is recognized by farmers to be important. Although
land preparation by machine may cost more than by animals, the
difference may be less than the value of the reduced work load for
the farmer and the flexibility in timing operations.
The cost-value structure changes continuously over time and differs
between farmers. Government policies can affect both sides of the
equation. In the Philippines, for example, rice price policies and
green revolution technology packages increased the value of flexibility
in timing operations while higher farm incomes probably allowed
farmers to consider alternative ways to reduce their work load.
Carabao costs began to increase in some areas partly due to increased
demand and partly as a result of a reduction in carabao numbers due
to serious floods. Local manufacturing of IRRI designed two-wheel
tractors reduced the cost of machine services. As a result, cost
differences became small enough to induce an increasing number of
farmers to purchase or rent small tractors.
In a survey of 150 rice farmers adopting small tractors in the
Philippines, 60 percent indicated they purchased tractors because of
the problems and nuisance value connected with carabao maintenance.
(Alviar, undated). A somewhat smaller number indicated that savings
in time was an important reason for purchasing a tractor. Farmers
apparently felt the savings in time was important for various reasons.
First, less time spent in land preparation allowed more time for other
productive activities. For the farmers sampled, fishing and duck
raising frequently provided additional family income. Second, farmers
apparently were aware that more rapid land preparation provided
greater flexibility in planting times. Being able to plant earlier
allows a wider choice of rice varieties, may help overcome labor
shortages and drying problems at harvest time, and can facilitate
multiple cropping. Finally, and in some cases perhaps most important,
the shift from carabao to small tractors probably increases a family's
leisure time as well as making some cultivation and harvesting tasks
easier. Where small tractors completely replace carabao the largest
saving in time is that previously devoted to caring for the carabao.
In summary, the survey results suggest that Philippine rice farmers
adopting small tractors anticipated very little, if any, direct yield
effect from mechanization. They adopted tractors because they believed
that tractors would make their life easier and would allow them to
alter the timing of field operations in ways that could reduce costs
or facilitate more intensive land utilization.
Similar results were reported in a 1967 survey in Thailand (Royal
Thai Government, 1969). Farmers who used tractors considered the
greater physical ease in accomplishing work and the time freed to
earn income from other sources as more important than the possible
cost or output advantages of tractors compared to other power
sources. The timing of farm operations appeared to be more important
on large farms which had a shortage of farm labor and draft animals
during peak farming seasons. The Thai study team reported that, in
general, the larger the farm the smaller the yield for most crops.
They suggested that increased mechanization could result in higher
yields on larger farms by enabling them to overcome labor shortage
problems. They did not, however, report whether large farms having
adequate tractor power did indeed have higher yields than large
farms with less tractor power.
The timing of agricultural operations frequently has been noted to
have an important effect on crop yields. In Kenya, corn planted to
take advantage of early rains reportedly yields over 50 percent higher
than that planted a month later. Early weeding was found to increase
corn yields by another 25 percent (Voss, 1974). In Turkey cereals
planted in October resulted in a 38 percent higher return than those
planted in November. In India it has been estimated that for each
day's delay in planting after the optimum period there is a one
percent reduction in wheat yields. Such evidence does not prove,
however, that farmers need tractors in order to plant or cultivate
their crop on time. Nor does the existence of technical opportunities
to increase yields by altering planting dates automatically assure
that farmers with tractors will take advantage of the opportunities.
The yield effect of tractorization has not been studied as carefully
in Africa or South America as in Asia. Abercrombie, after reviewing
the available information on mechanization in South America, con-
cluded that "better soil preparation, including such operations as
deep ploughing and subsoiling that are only possible with mechaniz-
ation, increases yields per hectare" (Abercrombie, 1972). This
judgement, however, is based on little empirical data and, if true,
would apply primarily to heavier soils. Statistical studies by other
researchers have not revealed a strong yield effect. William R. Cline,
for example, used regression analysis to separate the yield effect of
tractors and fertilizer for 117 rice farmers in southern Brazil
(Cline, 1970). His results showed that, at most, the use of tractors
increased yields by only two percent.
Clayton, attempting to determine the advantages and disadvantages of
mechanization in Africa, concluded that "increased yields can result
from better seed-bed preparation, particularly in the case of heavy
soils, and/or more timely operations, including planting" (Clayton,
1972). This observation reflects conventional wisdom more than
empirical data and fails to separate the "enabling" effect of
machines from the direct yield effect. It was qualified later in
Clayton's study with the statement that the improved yields observed
with tractor cultivation in Africa "have often been due to the
accompanying introduction of improved seeds, fertilizer and insect-
V. Mechanization and Intensity of Land Use
Another advantage claimed for mechanization is that it allows more
intensive land utilization. The evidence to support this claim is
inconclusive. Many, but not all, studies report more intensive
cropping on tractor farms compared to bullock farms. Few of the
studies however attempt to separate the cropping intensity impact
of tractors, or of tractors as enabling devices, from the other
factors that may account for a substantial part of the higher level
of cropping intensity sometimes observed on tractor farms. Farmers
using tractors, for example, may have greater access to credit, more
capable managers, or better water control systems.
Several comparisons of the intensity of land use on tractor farms vs
bullock farms are presented in Table IV. The 23 percent difference
in cropping intensity reported in the Philippines is in part accounted
for by better water control systems on the tractor farms included
in the study. The IBRD study on the consequences of farm tractors
in Pakistan by Mclnerney and Donaldson (1973) is one of the more
careful efforts to estimate the cropping intensity effect of tractors.
They found cropping intensity on tractor farms to be seven percent
greater than on bullock farms during the 1966-70 period.
The degree of cropping intensity may increase over time on tractor
farms. If so, a short run comparison of tractor farms and bullock
farms may underestimate the longer run impact of tractors in enabling
more intensive land utilization. There is very little data, however,
on changes in cropping intensity over time resulting from mechanization.
A research team reviewing the 1957-67 data on tractor use in Thailand
concluded that the availability of farm equipment did not significantly
effect farming cycles or the extent of multiple cropping (Royal Thai
Government, 1969). The study recognized that tractors enabled
earlier plowing but found that while some farmers did start plowing
a little earlier, this did not result in any significant change in
existing cropping patterns. Many farmers preferred to utilize their
tractor to delay land preparation until adequate rainfall was assured.
As a result they followed basically the same cropping pattern as
farmers using animals.
S.S. Johl of the Punjab Agricultural University, using 1967-70
data from India's Punjab region, found that a 79 percent increase
in tractor use led to a 52 percent reduction in bullock use, a
20 percent increase in labor inputs and a nine percent increase
in cropping intensity. His results are summarized in Table V.
Using eight hour days to measure the change in inputs; as tractor
use increased by approximately one day per acre, bullock use declined
by five days and labor use increased by ten days.
The studies surveyed in this section suggest that tractorization
alone probably does not increase cropping intensity by more than
ten percent on the average. The impact of other types of mechaniz-
ation on cropping intensity have not been investigated carefully.
The introduction of irrigation pumps and tube wells in India and
Pakistan is the possible exception. These appear to have resulted in
TABLE IV: CROPPING INTENSITY OF MECHANIZED vs NON-MECHANIZED FARMS
Tractor Bullock Tractor
Farms Farms to
Crop % % Bullock
(1) (2) (3)=(l)+(2)
Philippines (1964-65) 1/
Rice 200 162 1.23
Pakistan (1966-70) 2/
Various 119 111.5 1.07
India (1972) 3/
Zone I 159 162 0.98
Zone II 127 134 0.95
India (1971) 4/
Various 120 114 1.05
Note: *Cropping intensity is measured by the total number of hectares
planted per year divided by the number of arable hectares. Thus,
double cropping of all arable hectares results in a cropping
intensity index of 200 while an index of 162 indicates that 62%
of arable land was double cropped.
Source: 1/ Alviar, undated study. 4/ Donde, 1972.
2/ McInerney and Donaldson, 1973.
3/ Parthasarathy and Abraham, 1974.
a greater increase in cropping intensity than the introduction of
tractors. The introduction of grain harvesting and drying equipment
theoretically could enable more intensive cropping in some areas
but there is not enough field research to determine if the theoretical
opportunities are realized in practice.
VI. Mechanization and On-farm Employment
In microeconomic theory, output is viewed as a function of inputs which
frequently are assumed to be interchangeable. Rice production, for
example, can be viewed as a function of land, labor, and capital inputs.
In theory, if capital inputs are increased with land and labor held
constant, output should increase. Alternatively, if output and the
land input are held constant while capital is increased, the rational
producer would reduce labor inputs. This theoretical model frequently
is the starting point in a chain of reasoning which leads to the
conclusion that mechanization will increase unemployment. Various
approaches have been taken to support the alternative conclusion
that mechanization, in fact, increases employment. One approach is
to emphasis the potential off-farm employment effects of mechanization.
Here it is argued that labor inputs are required to build and main-
tain the machines and that these inputs may more than offset the
reduction in farm employment. A second approach is to treat
mechanization as a shift in the production function. Machines are
viewed as completely new inputs which embody the latest technology.
They enable farmers to increase yields, utilize land more intensely,
and increase labor productivity. The increased yields and more
intensive use of land are assumed to increase total employment while
at the same time output per unit of labor input is increased.
The data used to support either the "employment" or "unemployment"
conclusion is inconclusive. The major difficulty again is the
failure to separate the employment impact of mechanization from
that of other technological or institutional changes that may occur
simultaneously with mechanization. The reliability of much of the
employment data is subject to question. The estimates of labor
inputs required to complete the same farming operation vary widely
between countries as well as between studies in a given country.
Part of the differences could be explained if more information were
available on the type of land, weather conditions, and particular
cultivation methods. Nevertheless, when one study reports labor
requirements for land preparation to be five times greater than
those reported in another study for the same crop using essentially
the same cultivation practices, it appears advisable to use the data
The proposition that mechanization reduces the amount of labor
per 'unit of output is widely accepted but there are only general
estimates of rate of technical substitution of farm machinery for
labor. The available data on labor use per hectare, with and
without mechanization, could be used to estimate changes in labor
input per unit of output if it is assumed that mechanization has
no yield effect. Most estimates of labor use per hectare fail to
quantify accurately the degree of mechanization used. As a result,
it is impossible to estimate, from available data, the extent to
which particular machines or implement combinations actually sub-
stitute for labor per unit of output.
TABLE V: CHANGES OCCURRING AS TRACTOR USAGE
IN THE PUNJAB, INDIA
1/ Ratio of gross cropped area to size of
INCREASED IN FARMS
(hr/acre) Cropping 1/
Results of a study by S.S. Johl reported by Drew and Bondurant,
1972. Table adopted from Stout and Downing, 1974.
A number of rough estimates of the "labor replacement" impact of
machines have been developed, however, which provide "rules of
thumb" for estimating the employment impact of farm equipment in
For example, using farm management surveys, Billings and Singh
estimated the per acre labor replacements of different mechanical
devices used in the Punjab, India as follows:
1. Pumpsets at 1/4 the man-hours required with a Persian wheel,
2. Wheat threshers at 1/4 the man-hours as indigenous methods,
3. Tractors at 1/5 of man-hours required using bullocks,
4. Reapers at 1/5 the man-hours needed with traditional methods, and
5. Corn sellers at 1/7 the labor formerly required (Billings and
Using monthly estimates of labor replacement of various levels of
mechanization on an irrigated 10-acre farm in the Punjab, they found
that the introduction of high yielding grain varieties (HYV) produced
with conventional power sources in 1968-69 would have increased labor
use by six percent. HYV with pumpsets would increase labor require-
ments by only two percent while the use of HYV, pumpsets, wheat
threshers, cane crushers, corn sellers, tractors, and reapers would
reduce total labor requirements by nearly six percent. Projecting
the impact of mechanization to 1983-84, HYV with a high level of
mechanization resulted in a 17 percent net reduction in labor use
compared to using conventional farming methods with HYV. These results
are summarized in Table VI. They illustrate the substantial employment
impact of the HYV. Wheat threshers are estimated to cause the largest
reduction in employment in the short-run while the introduction of
tractors has the greatest long-run impact.
There appears to be wide agreement that the long-run effect of farm
machinery on agricultural employment is likely to be substantially
greater than the short-run effect. During the early stages of
mechanization only a few operations are mechanized. Water pumping,
land preparation, and grain threshing are frequently the first
operations mechanized. The employment impact of this mechanization
may not be readily apparent in the short-run because of other
technological changes such as the introduction of HYV taking place
simultaneously. Over the longer run, however, tractors begin to be
used for additional mechanization. The long run impact of farm
mechanization is clearly illustrated by the United States experience.
Labor requirements for wheat production in the United States fell
from 160 man-hours per hectare in 1830 to 6 in 1930 (Abercrombie, 1972).
For agriculture as a whole, labor input per hectare was reduced by
two-thirds between 1896 and 1953, while machinery input per hectare
rose more than sixfold (Kendrick, 1961).
TABLE VI: CHANGES IN THE DEMAND FOR LABOR AS A RESULT OF
MECHANIZATION IN INDIAN PUNJAB
Level Estimated Projected
of 1968-69 1983-84
Technology (%) (%)
Conventional 100 100
1. with HYV 106 113
2. with pumpsets 102 103
3. with wheat threshers 96 96
4. with cane crushers 95 94
5. with corn sellers 95 92
6. with tractors 94 86
7. with reapers n.a. 83
Net change 1/ -6 -17
1/ Conventional minus line 6 for 1968-69.
Conventional minus line 7 1983-84.
Source: Billings and Singh, 1970.
Estimates of the labor replacement impact of tractors vary widely.
McInerney and Donaldson (1973) estimated that the use of four wheel
tractors in Pakistan during 1966-70 reduced labor use per cultivated
hectare by some 40 percent, resulted in 4.2 tenant families replaced
per farm, and a net overall destruction of about five jobs per
tractor. Abercrombie, using data for Colombia, noted that the impact
of tractors on employment varies with farm size. He estimated that
as many as 18.9 workers could be displaced per tractor on the average
for farms in the 50 to 199 cultivated hectares range. For farms
with over 200 cultivared hectares, however, the substitution ratio
drops sharply to 2.3 workers per tractor (Abercrombie, 1972). For
Colombia overall, Abercrombie estimated that the introduction of one
tractor resulted in a reduction of average labor requirements for
major field crops of 5.7 man years. This compares with 4.1 man years
in Chile and 6.8 man years in Guatemala. The smaller reduction in
Chile than in the other two countries maybe accounted for by the
higher proportion of irrigated crops and greater use of animal
power in Chile compared to Guatemala where very little animal power
is used in traditional agriculture. Abercrombie's estimates are
based on labor requirements rather than actual use. Therefore,
they illustrate the potential impact of tractorization rather than
actual changes in rural employment during a particular time period.
Several researchers have noted that increased use of tractors appears
to reduce employment less on large farms. In some cases, large farms
may find it profitable to maintain more workers than needed most of
the time in order to have adequate labor supplies during peak employ-
ment periods. Other researchers report a larger labor replacement
effect for tractors used on large farms. What may happen is that
when tractors are first introduced on large farms, workers are pro-
vided other on-farm employment opportunities but that the nubmer of
new opportunities diminish rapidly with increased mechanization.
The employment impact of mechanization depends partly on the mixture
of crops grown. Differences in labor requirements for major crops
produced with and without mechanization in several Latin American
countries are presented in Table VII. Mechanization appears to
reduce labor requirements for potatoes by only 6 to 19 percent
compared to a 50 to 90 percent reduction for wheat. The overall
average for the five crops and four countries is approximately a
50 percent reduction in labor requirements with mechanization
(Abercrombie, 1972). Again, actual labor replacement may be less
than the potential estimated replacement. Nevertheless, the
potential reduction in labor use appears large enough to justify
the conclusion that mechanization generally will reduce employment
per hectare as well as per unit of output other things remaining
It has been suggested that this conclusion, while applicable to
large four wheel tractors, does not apply to small two wheel tractors
(Javed Hamid, 1973). Basically, the argument is that small two wheel
tractors replace animals not farm workers while large tractors pre-
sumably replace both. Farm survey data for the Philippines does not
support this conclusion. Based on a small sample of rice farmers in
Nueva Ecija, Philippines, Bautista, and Wickham (1974) estimated labor
inputs of primary and secondary tillage to be 7.4 man days per hectare
using carabao compared to 4.0 days for farmers using small two wheel
tractors. This represents a 46 percent reduction in labor use for
land preparation alone. The farmers surveyed recognized very little
yield difference after using tractors compared with the period when
carabaos were used. They did, however, observe less weed competition
on land prepared by tractor but there was no indication that this
lowered labor inputs for weeding. The labor replacement effect of
two wheel tractors was only slightly less than that of four wheel
TABLE VII: LABOR REQUIREMENTS, WITH AND WITHOUT MECHANIZATION, IN
SELECTED LATIN AMERICAN COUNTRIES
Corn Rice 1/ Beans Potatoes Wheat
(man-days per hectare except as noted)
a. Without Mechanization 60 48 70 75 26
b. With Mechanization 35 33 50 65 10
c. (b) as % of (a) 58 69 71 87 38
a. Traditional 49 71 62 125 32
b. Modern-non-mechanized 3/ 78 93 82 193 63
c. (b) as % of (a) 159 131 132 154 197
d. Modern Mechanized 30 36 18 156 7
e. (d) as % of (b) 38 39 22 81 11
a. Human Energy Only 56 103 57 162 103
b. Human and Mechanical
Energy 44 54 44 153 47
c. (b) as % of (a) 79 52 77 94 46
a. Without Mechanization 48 85 -- 16
b. Semi-mechanized 8 26 -- 6 4/
c. (b) as % of (a) 17 30 38
1/ Irrigated rice.
2/ Data for irrigated crops.
3/ Theoretical situation of improved agriculture without mechan-
4/ Fully mechanized.
tractors. Their estimates of the changes in labor use resulting
from the rental of four wheel tractors are presented in Table VIII.
Total labor use was reduced by 52 percent. Family labor inputs
declined by 54 percent while hired and exchange labor inputs fell
38 percent as a result of renting tractors.
The impact of small hand tractors on employment also depends partly
on the type of crop grown. Banta used field experiments to compare
labor use with three levels of power inputs; hand labor, carabao,
and two-wheel tractors. His results are shown in Table IX. For rice
production the use of carabao required only 59 percent of the labor
inputs of hand operations while small tractors required only 30 per-
cent as much labor as hand cultivation and only 51 percent the labor
inputs required when using a carabao. In the production of sweet
potatoes the use carabao had an even greater effect on labor use
than did a hand tractor. When cowpeas and corn were intercropped,
carabao required only 46 percent of the labor input of hand operation
while two wheel tractors required only 34 percent as much labor as
hand cultivation methods. The substitution of animal power for
land labor appears to result in an even greater reduction of small
tractors for animal power.
Additional data on the employment impact of small tillers used for
rice production in the Philippines is presented in Table X. The
labor input used in land preparation is nearly 73 hours per hectare
when carabao are used for both plowing and harrowing (technique T5)
and only 34 hours per hectare when small tillers are used for both
operations (technique T3). Small tillers reduce labor inputs by over
25 percent when used only harrowing and by over 50 percent when used
for both plowing and harrowing.
Estimates of a fifty percent reduction in labor requirements are also
reported in other studies. Bose and Clark (1969) using field survey
data for Pakistan reported that, "(in) interviewing farmers in the
Punjab who have mechanized, we received a remarkedly consistent res-
ponse that the labor force per acre had been reduced about 50 percent
from the pre-mechanization period." McFarguhar and Hall (1970) using
data on cotton production in Uganda reported a 57 percent reduction
in man-hours per acre for mechanized vs hand-hoe cultivation.
These estimates provide little information on the changes in crop
production patterns or in the composition of the labor inputs
which frequently take place simultaneously with mechanization. Some
studies report a reduction in family labor inputs and an increase in
use of hired labor. Others report reductions in both family and hired
labor inputs and in some cases family labor inputs reportedly increase
while hired labor is reduced. In some countries, machine rental
services are widely available, in other cases farmers own their own
TABLE VIII: IMPACT OF USE OF FOUR WHEEL TRACTORS, NUEVA ECIJA,
Ha 1/ Percentage
1. Labor Use Before Renting Tractors 7.3
2. Labor Use After Renting Tractors 3.5
3. (2) + (1) 48
4. Family Labor Used Before Renting Tractors 9.6
5. Family Labor Used After Renting Tractors 4.4
6. (5) + (4) 46
7. Hired & Exchange Labor Before Renting
8. Hired & Exchange Labor After Renting
9. (8) t (4) 62
Work Done With Labor Saved by Using
Tractor Services (% of farms)
a. Fix and clean bunds 57
b. Cleaning before Final Harrowing 27
c. Work on other farms for pay 1
d. Non-farm work 15
I/ Unweighted averages of wet and dry season data for farmers
renting four-wheel tractors.
Source: Bautista and Wickham, 1974.
TABLE IX: LABOR USE WITH THREE POWER SOURCES FOR SELECTED CROPS
IN THE PHILIPPINES, 1973
Labor Carabao Tractor
(1) (2) (3)
(Hours per hectare)
Hand Tractor vs
Carabao Tractor Carabao
(2)%(1) (3)+(1) (3)+(2)
(M) (%) (M)
Cowpea & Corn
1/ Labor time in hours per hectare.
2/ Carabao time in hours per hectare.
3/ Tractor time in hours per hectare.
Source: Gordon R. Banta, 1973.
TABLE X: LABOR INPUTS USED IN LAND PREPARATION, NUEVA ECIJA,
Land Labor Input in (hr/ha) As Mean
Preparation Percent Yield
Technique 1/ Plow Harrow Total of T5 (MT/ha)
Tl 5.09 40.02 45.11 62 3.97
T2 7.60 40.59 48.19 66 4.00
T3 13.38 20.56 33.94 47 4.14
T4 33.99 19.84 53.83 74 4.01
T5 32.28 40.34 72.62 100 3.94
1/ Data apply to medium hardpan soil conditions during the 1973
T1 Large tractor plowing and carabao harrow.
T2 Large tiller plowing and carabao harrow.
T3 Small tiller both plow and harrow.
T4 Carabao plow and small tiller harrow.
T5 Carabao both plow and harrow.
Source: Orcino and Duff, July, 1974.
tractors and apparently are reluctant to rent them or undertake custom
work. These differences appear to be the result of differences in
institutional structures. The introduction of tube wells and
irrigation pumps frequently increases the opportunities for multiple
cropping and gradually results in changes in cropping patterns.
Where tractors actually replace bullocks, some land formerly devoted
to fodder crops may be devoted to other crops. In a few cases, the
demand for various crops appears to shift rapidly either just prior
to significant changes in the rate of mechanization or during the
early stages of mechanization. This results in a supply response that
involves changing the composition of production. Many of the production
changes probably would occur with no changes in mechanization but
mechanization may enable the changes to take place more rapidly either
for technical reasons (increased water availability, for example) or
by inducing changes in relative factor prices.
The factors causing the changes which take place simultaneous with
mechanization are difficult to identify and in many cases nearly
impossible to quantify. Different opinions about the direction of
the cause-effect relationships can lead to substantial differences
in conclusions concerning the "effects" or the "value" of mechaniz-
ation. This can be illustrated using the data in Table XI on the
influences of tractors on employment in India. These data are based
on a study of 76 farms having 79 tractors in the Dhulia district in
Hamharashtra, India (Donde, 1972). The increases in employment per
hectare ranging from 10 to 32 percent reportedly are "the result of
tractor mechanization" (Stout and Downing, 1974). Much of the employ-
ment increases could, however, be attributed to the changes in farm
structures which occurred simultaneously with mechanization. These
1. 50 percent of the wasteland was reclaimed,
2. net irrigated area increased by 24 percent,
3. the double cropped area increased from 14 percent of the
net sown area before tractors to 20 percent after,
4. the cropping pattern changed towards more intensively grown
crops after tractors were introduced.
The question is, "Could these changes have taken place (or would they
have taken place) without the introduction of tractors?" Were tractors
essential to reclaiming the wasteland? Did tractors cause a change
in cropping patterns or did the changes reflect changes in relative
prices and costs? Were tractors needed to increase the irrigated
area? Was the increase in area double cropped the result of tractors
or due to the increased availability of irrigation water? If one
concludes that tractors "caused" the changes in farm structures, then
it follows that the introduction of tractors did in fact "result in
an increase in labor inputs." Such a conclusion certainly is debatable.
It appears that much of the data presented to support the argument
that mechanization leads to an increase in farm employment is mis-
interpreted and reflects a failure to separate the employment effect
of machinery from that of other technological changes occurring
simultaneously with mechanization. Theoretically, large tractors
(30 to 90 hp) could reduce labor inputs by at least 80 percent and
small two-wheel tractors (7 to 15 hp) probably by at least 60 percent
compared to use of bullocks. In practice, it appears that in grain
production, large tractors probably reduce labor inputs per hectare
by 40 to 50 percent and small tractors by 30 to 40 percent. The
introduction of tube wells with irrigation pumps appears to be the
only case in which mechanization may have a significant positive
effect on employment either through higher yields or more intensive
VII. Mechanization and Off-farm Employment
An assessment of the total employment effects of mechanization must
take account of the employment generated by the manufacture, distri-
bution, maintenance, and repair of farm equipment. It is often
argued that such off-farm employment goes a long way towards offsetting
any displacement of agricultural labor possibly caused by mechanization.
Estimates of the off-farm employment effects of mechanization are even
less precise than the on-farm estimates. Nevertheless, the limited
available evidence suggests that the off-farm employment resulting
from increased mechanization is rather small relative to the loss of
Abercrombie's estimates of employment in agricultural machinery
manufacture, distribution, maintenance, and repair in LAFTA countries
are presented in Table XII (Abercrombie, 1972). The estimates admittedly
are very rough but, if anything, probably overestimate total employ-
ment. For the eleven LAFTA countries in 1968 total employment in
agricultural machinery and related activities was less than 150,000.
Although these jobs are at a much higher productivity and income
levels than agricultural jobs, their number is almost insignificant
in the over-all employment situation. They represent only 0.2
percent of total employment and approximately 0.5 percent of the
number of persons employed in agriculture.
TABLE XI: INFLUENCE OF TRACTORS ON EMPLOYMENT IN INDIA
Farms Farms Difference
(1) (2) (1)-(2) (1)t(2)
(man days/ha.) %
Casual labor employment 91 69 +22 132
Employment on tube-well farms 104 88 +16 118
Employment on canal-irrigated farms 51 46 + 5 110
Source: Donde, 1972 as reported by Stout and Downing, 1974.
ESTIMATED EMPLOYMENT IN AGRICULTURAL MACHINERY
MANUFACTURE, DISTRIBUTION, MAINTENANCE, AND
REPAIR IN LAFTA COUNTRIES (1968)
Domestic Manufacture Inter-
Tractors Other Products
(1000 persons employed)
a. Based on estimate of 34 percent purchases on the internal market
b. Total includes distribution, maintenance, and repair of agric-
ultural machinery including imports.
c. Applying the average of the Argentina and Mexico ratios to
d. Applying the Argentina ratio to total sales.
e. Rough estimate.
Source: Abercrombie, 1972.
Using data for Brazil and Argentina, Abercrombie estimated that an
investment of up to US $10,000 is required to create one job in
tractor manufacturing which will produce two to five tractors each
year. With some additional investment, one further job may be
associated with the distribution of these tractors and the main-
tenance and repair of the existing stock. Another one-third of a
job may result from backward linked industries. This is only 2.3
non-agricultural jobs resulting from the production of two to five
tractors annually, or one such job from the production of 0.9 to
2.2 tractors. Each tractor has the potential to reduce farm employ-
ment by five man years during each year of operation. Assuming
that tractors last ten years, each off-farm job generated by tractor
production potentially could displace from 4.5 to 11.0 agricultural
workers annually over the ten year tractor lifetime. In practice,
each tractor probably does not reduce employment initially by more
than half of its potential. Even so, during the early stages of
mechanization, each man-year of non-agricultural employment associated
with mechanization would reduce on-farm employment by 20 to 50 man-
years over a ten year period. In Abercrombie's judgement, the lower
limit on the replacement ratio for Latin America is approximately 1
Similar conclusions apply to the local manufacture of small two-wheel
tractors in the Philippines. Local manufacturers of small two-wheel
tractors (7.5 hp) estimate that it requires from 5 to 20 man-days
to produce and assemble a small tractor depending on the type of
equipment available to the plant and the size of the production run.
Using an average of 12.5 days per small tractor assembled, a pro-
duction worker would produce approximately 20 small tractors annually.
Backward linkages are relatively small in this case because a large
part of the material for the tractors is imported. Labor require-
ments to distribute and service the tractors probably are not more
than two man-years per 20 tractors. In this case, each man-year of
non-agricultural employment results in 6.7 small tractors which
normally would remain in production for at least 5 to 7 years. The
on-farm employment effects of these tractors can be estimated using
the data from two IRRI studies. Orcino, using a small survey of 45
hadn tractor users in Laguana (Philippines), estimated the average
7.5 hp tractor was used approximately 600 hours annually in 1970
(Orcino, 1972). Banta's field experiment data in Table IX indicate
that 25 hours of tractor time are required to produce a hectare of
rice and that the use of hand tractors for rice production reduced
labor requirements by 190 hours per hectare (Banta, 1973). Using
these estimates, the hand tractor used 600 hours could cultivate
24 hectares of rice and would reduce labor requirements by 4560
hours. Assuming 225 eight hour work days annually, each small
tractor would reduce labor requirements by 2.5 man-years annually
or approximately 12 to 18 man-years over a five to seven year period.
If 6.7 tractors are produced and maintained per man-year of off-
farm employment then the result is a total reduction of 80 to 120
man-years of on-farm employment in rice production per man-year of
employment with cowpeas, each year of employment off-farm to produce
small tractors would reduce on-farm employment by 35 tp 50 man-
years over a five to seven year period.
These results, though based on very different data than used by
Abercrombie, lead to very similar conclusions; as a minimum, a
man-year of non-agricultural employment in the production, sales
and service of tractors (whether large or small) reduces on-farm
employment by 20 to 30 man-years.
The estimates for both South American and the Philippines are very
rough. Nevertheless, even if the off-farm employment per tractor
was doubled and the on-farm reduction in employment reduced by 50
percent, the off-farm employment created would still be far short
of off-setting the reduction in on-farm employment.
VIII. Approaches to Increasing Power Available to Agriculture
By nearly any standard of comparison, the amount of power (human,
animal and machine) available to agriculture in developing nations
is low. More intensive cultivation will require increases in the
availability of agricultural power and/or more intensive and
effective utilization of available power sources. The available
data suggest that agricultural mechanization is increasing throughout
the world, although relatively slowly in many countries and with wide
regional differences in the rate of mechanization within countries.
The high rates of rural unemployment and underemployment reported
in many LDC's suggest that available manpower is not fully utilized.
Whether this is because the rate of mechanization has been too fast
or too slow has never been answered adequately for any particular
country and probably cannot be answered in general.
Population growth in most countries increases human power availability
by two to three percent annually. This adds very little to net
power availability. A man develops approximately 0.1 HP using food
as his energy source. Where unemployment and underemployment rates
are high, the additional manpower is not fully utilized. When
utilized with traditional technology, man consumes 10 to 20 percent
of his food energy output and uses much of the the remainder to main-
tain his family. Furthermore, in most countries the marginal physical
productivity per available human horsepower is likely to decrease as
the man/land ratio increases and poorer quality land is brought into
production. Thus while increased manpower may result in more intensive
cultivation, it is unlikely to result in any substantial increases
in per capital food production.
In some areas, particularly in parts of Africa, increases in the number
of draft animals may be an appropriate way to increase power available
to agriculture. Draft animals (cows and bullocks) can provide up to
0.6 HP and can utilize energy from grasses and food by-products not
used for human consumption. In traditional agricultural systems land
held in fallow frequently provides the feed requirements of animals.
When only small amounts of land are held fallow, grain by-products
and grasses from field margins provide most of the required livestock
feed. The feed costs for draft animals used in traditional agricultural
systems are fairly low but the opportunity costs of fallow land used
for animal feeds may increase substantially with the introduction of
modern technology. Where unemployment and underemployment rates are
high, the immediate labor costs of caring for draft animals also are
relatively low. Draft animals are difficult to maintain, however, in
locations where water is scarce, annual rainfall is extremely variable,
or livestock diseases are common.
Under favorable conditions available animal power could be doubled
every ten years or so. Where most arable land is already in production
and human population is doubling every 25 to 30 years, increasing
animal numbers does not, however, appear to be a viable long-run
approach to increasing agricultural power. Furthermore, there is a
point at which the maintenance cost of keeping high levels of animal
power per hectare becomes prohibitive. The introduction of modern
agricultural systems increases the opportunity cost of the land and
labor required to maintain animals and lowers the relative cost of
fossil fuel power sources. Thus programs to introduce modern pro-
duction technology and to increase draft animals numbers become
increasingly non-complementary as the availability of unused arable
land declines. Programs to increase draft animal numbers, therefore,
should be viewed as intermediate term programs which may be appro-
priate in situations where available agricultural power is below
approximately 0.2 HP per hectare, unused arable land is relatively
abundant, water supplies are adequate and reliable, and livestock
diseases are not a serious problem.
In those countries where draft animals are used widely, programs to
up-grade the quality of animals, improve livestock management, and
develop more effective animal power implements probably would result
in better utilization of available animal power. There exists very
little information,however, by which to judge the potential of such
programs. It seems unlikely that they would increase the rate of
return on the average farmer's investment in draft animals by more
than ten to twenty percent. Furthermore, such programs would have
to reach a large number of small farmers in order to have a signi-
ficant total impact. Thus, animal power improvement programs pro-
bably would be of greatest value when included as part of a more
general technology improvement package designed for small farmers
using draft animals.
Increased mechanization appears to be the only viable means of
significantly increasing available agricultural power in most
countries over the long run. It seems likely that liquid fossil
fuels will continue to be the lowest cost energy source for machines
providing mobile power; at least for the next twenty to thirty years.
Conservation of liquid fossil fuels will require increased use of
solar energy, gravity systems, and wind power as well as more
complete and efficient utilization of agricultural by-products.
Non-fossil fuel energy sources can be most readily used for
stationary power units. Rice hull burners for grain drying and
windmills to pump irrigation water are examples of stationary power
units utilizing non-fossil fuel energy.
Most government programs to encourage farm mechanization during the
1960's emphasized the introduction of tractors. These programs,
which frequently were supported by international lending agencies,
have received more thorough evaluation than most other types of
mechanization programs. Many of the tractor mechanization programs
have been judged to be failures in the sense that estimated social
costs exceeded social benefits. The tractorization studies have
provided estimates of labor replacement but little empirical data
on actual unemployment created by the introduction of tractors. They
do, however, indicate that large farmers have usually benefited more
than small farmers from government programs to subsidize tractor
The conclusion that tractorization programs have seldom increased
social welfare has not led to a rejection of mechanization programs
in general. It has, however, led to various proposals which place
more emphasis on minimizing unemployment effects and improving income
One approach is to promote selective mechanization. That is, to
mechanize those operations which will reduce costs, have the least
effect on employment, and the greatest effect on output. Selective
mechanization hopefully would improve working and living conditions
in rural areas thus making rural living more attractive and reducing
migration to urban areas (Stout and Downing, 1974). The emphasis
is on increased productivity of labor while maintaining employment
per unit of land area. Those operations where accuracy and timing
are important for increased yields, for example, may be selected
for mechanization provided they reduce per unit production costs
and increase labor productivity without substantially reducing
A second approach, sometimes referred to as fractional mechanization,
focuses almost exclusively on the small farmer. It is essentially a
"small is beautiful" approach. Tractors are viewed as "lumpy inputs"
which frequently are inappropriate for small farmers because their
services are difficult to allocate to small land units. The alter-
native is to develop and introduce small machines which will replace
animals rather than people. Small two-wheel tractors are an example
A third approach is to promote appropriate mechanical technologies.
This approach views the mechanization of agriculture as a continuous
and inevitable process in economic development but one whose speed
and direction can be altered by public policies and programs. The
criteria used to define what is "appropriate" depend on a nation's
goals. If a nation's development program emphasizes equity, employ-
ment and production, then particular agricultural implements would
be considered appropriate if:
1. their use is profitable for a relatively large proportion of
the nation's farmers,
2. they can, to a large extent, be produced locally,
3. they make the maximum possible use of domestically available
resources such as energy, materials, and labor skills, and
4. they provide more effective and timely operations which
enable farmers to increase output either through increased yields
or more intensive use of their land resources.
Obviously, what is "appropriate" for one country may not be appropriate
for another, or even for the same country at a later stage of develop-
ment. Less obvious, but perhaps more important, is the implication
that appropriate mechanical technologies, or at least the most appro-
priate ones, may not yet be available to many countries.
The appropriate mechanical technologies approach appears to have
wide support. It reflects a sensitivity to unemployment and income
distribution problems, an awareness of the technical limitations of
small machines, and focuses on a few key aspects of the social and
technical infrastructure required for successful mechanization
Abercrombie, K.C., "Agricultural Mechanization and Employment in
Latin America," International Labour Review, CVI: 11-45,
Advisory Group, Program Proposal for the Agricultural Engineering
Department of the Internationa Rice Research Institute, Los
Banos, Philippines, IRRI, March 1974.
Agrawal, R.C., D. Rao and M. Singh, "A Study of Factors Affecting
the Demand for Rural Labour in Agriculture," Indian Journal
of Agricultural Economics, XXV: 60-63, 1970.
Ahmad, Bashir, A Study into Economic Comparison of Bullock Farming
versus Tractor Farming in Lyallpur District, M.Sc. Thesis,
Faculty of Agricultural Economics and Rural Sociology, West
Pakistan Agricultural University, Lyallpur, Pakistan, 1970.
Ahmad, Bashir, Farm Mechanization and Agricultural Development:
A Case Study of the Pakistan Punjab, Unpublished Ph.D.
dissertation, (Michigan State University, 1972).
Aktan, Resat, "Mechanization of Agriculture in Turkey." Land
Economics, XXXIII: 273-284, November 1957.
Alviar, Nelly, A Study of Tractor and Carabao Cultivated Farms in
Laguna, Los Banos, Philippines, IRRI, undated.
Anker, D.L.W., "Some Effects of Farm Mechanization," International
Labour Review, LXXI: 236-265, March 1955.
Banta, Gordon R., Comparison of Power Sources in Multiple Cropping,
Los Banos, Philippines, IRRI, August 1973.
Barker, Randoloph, W.H. Meyers, C.M. Crisostomo, and B. Duff,
"Employment and Technological Change in Philippine Agriculture,"
International Labour Review, 106: 3-31, August-September 1972.
Bautista, Fe and T. Wickham, The Tractor and the Carabao: A Socio-
Economic Study of the Choice of Power Source for Land Preparation
In Nueva Ecija, Los Banos, Philippines, IRRI, July 1974.
Billings, Martin H. and A. Singh, Farm Mechanization and the Green
Revolution, 1968-1984: The Punjab Case, New Delhi, India, USAID,
Billings M. and Arjan Singh, "Mechanization and Rural Development with
Some Implications for Rural Income Distribution," Economic and
Political Weekly, V26: A61-A72, June 1970.
Bose, Swadesh R. and Edwin H. Clark II, "Some Basic Considerations
on Agricultural Mechanization in West Pakistan," The Pakistan
Development Review, IX: 273-308, Autumn 1969.
Bradfield, Richard, Mechanized Maximum Cropping Systems for the
Small Farms of the Rice Belt of Tropical Asia, Los Banos,
Philippines, IRRI, March 1971.
Cervinka, Vaclav, "Tractor and Implement Efficiency on Tropical
Farms," World Crops, XXIII: 136-138, May-June 1971.
Chancellor, W.J., Survey of Tractor Contractor Operations in Thailand
and Malaysia, Davis, California, University of California, August
Chancellor, W.J., Social and Institutional Factors in Mechanization
of Agriculture in Thailand and Malaysia by Tractor Hire Services,
Davis, California, University of California, 1972.
Clayton, Eric S., "Mechanization and Employment in East African
Agriculture," International Labour Review, CV: 309-334, April
Cline, William R., Economic Consequences of a Land Reform in Brazil,
Amsterdam, North Holland Publishing Company, p. 188, 1970.
Coleman, P.C., "Wet Padi Mechanization Investigations in Province
Welleskey During the 1950-51 Season, Malayan Agricultural Journal,
Crisostomo, C.M., W.H. Meyers, T.B. Paris, Jr., B. Duff, and R. Barker,
"The New Rice Technology and Labor Absorption in Philippine
Agriculture," Malayan Econ. Rev., XVI: 117-158, October 1971.
Dalrymple, D.G., "Mechanization and Social Welfare: A Belated Comment,"
American Journal of Agricultural Economics, LIV: 537, August 1972.
Dalton, G.E. and J. Enikwaw, "Costs and Benefits from Operating Wheeled
Tractors in the North of Ghana," Accra, Ghana, Department of
Agricultural Economics, University of Ghana, July 1971.
Davies, Cornelius, Consideration and Procedures for the Successful
Introduction of Farm Mechanization, Rome, FAO, FAO Agricultural
Development Paper 44, 1964.
Davis, Ron E., Machinery for Rice, Washington, D.C., USAID, November 1972.
Davis, Ron E. and Judson M. Harper, Utilization and Marketing Review
of Selected Countries Evaluation and Recommendations for Agri-
cultural Equipment Development Research for Tropical Rice
Cultivation Project AID/csd-2541, Washington, D.C., USAID, Sept. 1973.
Dawlaty, Khairullah, "The Effects of Tractors on Farm Output, Income
and Employment During the Initial Stages of Farm Mechanization in
Afghanistan," Employment and Unemployment Problems of the Near
East and South Asia, R.G. Ridker and H. Lubell, ed., Vol. II,
Delhi: Vikas Publications, 1971.
Department of Agriculture, Federation of Malaya, Investigations into
the Mechanical Cultivation of Padi in Malaya, Mechanization Series
No. 1 Ministry of Agriculture, Swettenham Road, Kuala Lumpur,
Desai, D.K. and C. Gopinath, Impact of Tractorization on Productivity
and Employment (Gujarat State), Centre for Management in Agriculture,
Indian Institute of Management, Ahmedabad, India, 1973.
de Wilde, John C., "The Manpower and Employment Aspects of Selected
Experiences of Agricultural Development in Tropical Africa,"
International Labour Review : 104 (5), 1971.
Dilmus, D. James, The Economic Feasibility of Employing Used Machinery
in Less Developed Countries, Unpublished Ph.D. dissertation,
Michigan State University, 1970.
Donaldson, G.F., Operational Problems in the Introduction of Power
Tiller to Farms in East Pakistan, Washington, D.C., IBRD,
Donaldson, G.F. and J.P. Mclnerney, "Changing Machinery Technology
and Agricultural Adjustment," Am. J. Agr. Econ., 55: 829-839,
Donde, W.B., Impact of Farm Mechanization on Employment, Natiohal
Commission on Agriculture, India, 1972.
Drew, L.D. and B.L. Bondurant, "Effects of Land Holdings Size on
Agricultural Mechanization in India," ASAE Paper No. 72-683, 1972.
Duff, Bart and Ida Estioko, Establishing Design Criteria for Improved
Rice Milling Technologies, Los Banos, Philippines, IRRI, August 1972.
Eicher, Carl, Thomas Zalla, James Kocher, and Fred Winch, Employment
Generation in African Agriculture, Research Report No. 9, East
Lansing, Michigan, Institute of International Agriculture,
Michigan State University, 1970.
Ellis, G., Man or Machine: Beast or Burden: A Case Study of the
Economics of Agricultural Mechanization in Ada District, Ethiopia,
Unpublished Ph.D. dissertation, University of Tennessee, 1972.
Esmay, M.L. and L.W. Faidley, Agricultural Mechanization and Labor
Utilization in Asia, East Lansing, Michigan, Michigan State Univ-
ersity, June 1972.
Estioko, Ida, Farm Mechanization in the Philippines and the Inter-
national Bank for Reconstruction and Development Loan Program,
Penang, Malaysia, A/D/C Seminar on Farm Mechanization in South-
east Asia, December 1972.
Evenson, R.E., "Consequences of the Green Revolution," New Haven, Yale
University, Unpublished mimeo, 1973.
FAO, Agricultural Employment in Developing Countries, Agricultural
Planning Studies, No. 16, Rome, FAO 1973.
FAO, Final Report of the Agricultural Engineering Preliminary Project
Formulation Mission, Rome, March 1973.
FAO, The Role and Potentiality of Mechanization in Agricultural
Production in Africa, Rose Hill, Maurituis, Eighth FAO Conference
for Africa, August 1974.
Fisk, E.M., "The Mechanization of Agricultural Small Holdings,"
Malaysia Economic Review: 6(2), 1961.
Garst, R.B., The Spatial Diffusion of Agricultural Innovation in
Kisii District, Kenya, Unpublished Ph.D. dissertation, Michigan
State University, 1972.
Gayoso, Antonio, The Impact of Changing Technologies: Mechanization
and Employment, A Preliminary Review, Washington, D.C., USAID,
Gemmill, Gordon and C.K. Eicher, A Framework for Research on the
Economics of Farm Mechanization in Developing Countries, East
Lansing, Michigan, Michigan State University, April 1973.
Gemmill, Gordon and Carl Eicher, The Economics of Farm Mechanization
and Processing in Developing Countries, New York, Agricultural
Development Council, December 1973.
Giles, G.W., Agricultural Power and Equipment: The World Food Problem,
A report of the Consultants' Science Advisory Council, Vol. III,
Washington, D.C., 1967.
Giles, G.W., Agricultural Machinery Manufacture Feasibility Study
Mission, Iran, Rome, FAO, November 1972.
Gill, Muhammad Shaffi, "Economics of Farm Mechanization," West Pakistan
Journal of Agricultural Research, Lahore, West Pakistan, 1961.
Gotsch, Carl H., Agricultural Mechanization in the Punjab: Some
Comparative Observations from India and Pakistan, Working Paper
#11, AID/NESA 527, May 1973.
Gotsch, Carl H., "Tractor Mechanization and Rural Development in
Pakistan," Mechanization and Employment in Agriculture, ILO,
Guino, Ricardo A. and W.H. Meyers, The Effect of the New Rice
Technology on Farm Employment and Mechanization, Los Banos,
Philippines, IRRI, December 1971.
Hall, Malcolm, "Mechanization in East African Agriculture," Agricultural
Planning in East Africa, G.K. Helleiner, ed., Nairobi, East African
Publishing House, 1968.
Hamid, Javed, Agriculture Mechanization: A Case for Fractional
Technology, New York, Agricultural Development Council, September 1973.
Harrington, Roy E. et. al., Agricultural Mechanization in India, New
Delhi, India, The Ford Foundation, October 1972.
Hawkins, J.C., "From Manpower to Machines: Mechanization for Small
Farm in Under-developed Countries," World Crops, Vol. 13: 291-294,
August 1961; and Vol. 14: 351-355, September 1962.
Hu, C.H., Chuen-Bang Young, and Ten Kiau Huang, "Experimental Study
on the Comparison Between Conventional and Mechanical Methods of
Rice Cultivation in Taiwan," Bulletin of the Agronomy Club of
the College of Agriculture, Vol. 4, No. 1. Taiwan Provincial
Chung-Hsing University, Taichung, Taiwan, China, 1971.
ILO, Mechanization and Employment in Agriculture, Geneva, International
Labour Office, 1973.
International Rice Research Institute, Annual Report 1967, Manila,
Inkai, I., "Farm Mechanization, Output and Labour Input: A Case Study
in Thailand," International Labour Review, CI: 453-473, May 1970.
IRRI, Report of the Advisory Group Meeting on Rice Post Harvest Problems,
Los Banos, Philippines, IRRI, April 1974.
Jasny, N., "Tractor Versus Horse as a Source of Farm Power,"
American Economics Review, IIV: 708-723, December 1935.
Johl, S.S. "Mechanization, Labor Use and Productivity in Indian
Agriculture," Economics and Sociology Occasional Paper No. 23.
Department of Agricultural Economics and Rural Sociology, Columbus,
Ohio State University, 1970.
Johnston, B.F. and J. Cownie, "The Seed-Fertilizer Revolution and
Labor Force Absorption," American Economic Review, LIX: 569-582,
Johnson, Loyd, Power Requirements in Rice Production, Los Banos,
Philippines, IRRI, August 1963.
Johnson, Loyd, Mechanization to Improve the Standard of Living in
Tropical Rice-Growing Areas, New Orleans, La., Winter Meeting of
American Society of Agricultural Engineers, December 1964.
Johnson, Loyd, "Is Mechanization Possible Where Rice Is King?"
Agricultural Engineeing, 47: 132-135, March 1966.
Johnson, Stanley S., Performance and Economics of Use of Small
Equipment in Tropical Monsoon Countries: The Case of the Philippines,
Kyoto, Japan, Annuall Meeting of the Japanese Society of Agricultural
Machinery, April 1968.
Johnson, Stanley S., "Terminal Report on the General Engineering and
Economic Research Portion of Contract No. AID/csd-834 for Research
on Farm and Equipment Power Requirements for the Production of
Rice and Associated Food Crops in the Far East and South Asia,"
Los Banos, Philippines, IRRI, 1969.
Jones, E.H., "Busoga Farms, Bunya," in J.L. Joy (ed.): Symposium on
Mechanical Cultivation in Uganda, Kampala, Uganda Argus, Ltd., 1960.
Kaneda, Hiromitsu and Hohammed Ghaffar, "Output Effects of Tubewells
on the Agriculture of the Punjab: Some Empirical Results," The
Pakistan Development Review, X: 68-87, Spring 1970.
Karpat, K.H., "Social Effects of Farm Mechanization in Turkish Villages,"
Social Research XXVIII: 83-103, Spring 1960.
Kendrick, John W., Productivity Trends in the United States, Princeton,
for the National Bureau of Economic Research, Princeton University
Ker, A.D., "The Development of Improved Farming Systems for the Central
Plains Area of Uganda," Makerere University College, RDR paper 106,
Khan, Amir U., Appropriate Technologies: Do We Transfer, Adapt, or
Develop?, Los Banos, Philippines, IRRI, November 1973.
Khan, Amir and B. Duff, Development of Agricultural Mechanization
Technologies at the International Rice Research Institute,
Los Banos, Philippines, IRRI, January 1973.
Kishida, Yoshikuni, Connection Between Agricultural Productivity and
Tractor Power, Japan, Farm Machinery Industrial Research Corp.,
Kline, C.K., D.A.G. Green, Roy L. Davahue, and B.A. Stout, Agricultural
Mechanization in Equatorial Africa, Research Report No. 6, East
Lansing, Michigan, Institute of International Agriculture, Michigan
State University, 1969.
Kolawale, Michael Lyiola, An Economic Study of Tractor Contracting
Operations in Western Nigeria, Unpublished Ph.D. dissertation,
Cornell University, 1972.
Lindman, Russell, "The Tractor Factor: Agricultural Mechanization in
Peru," Public and International Affairs, Princeton, New Jersey, 1968.
Mangahas, Mahar, W.H. Meyers, and R. Barker, Labor Absorption in
Philippine Agriculture, Paris, France, OECD, 1972.
Mathur, P.N. and K.W. Kapp, "The Transition from a Bullock to a Tractor
Economy in India: Some Indirect Effects and Benefits,"
Wellwirtschaftliches Archir., Hamburg, LXXVII: 333-350, 1961.
McFarguhar, A.A.M. and M. Hall, "Mechanization and Agricultural
Development: No Miracle In Africa," Options Mediterraneennes,
Paris, December 1970.
Mclnerney, John P. and Graham F. Donaldson, The Consequences of Farm
Tractors in Pakistan, Washington, D.C., Development Economics
Department; International Bank for Reconstruction and Development, 1973.
Meij, J.L., Mechanization in Agriculture, Amsterdam, North-Holland
Publishing Company, 1960.
Mettrick, H., "Mechanization of Peasant Agriculture in East Africa,"
Change in Agriculture, A.H. Bunting, ed., New York, Praeger, 1970.
Miller, R.F., One Hundred Thousand Tractors: The MTS and the Develop-
ment of Controls in Soviet Agriculture, Cambridge, Harvard University
Mukumoto, Tsutomu, Changes and Mechanization in Agriculture in Japan,
Japan, Farm Machinery Industrial Research Corp., (undated).
National Council of Applied Economic Research, Impact of Mechanization
in Agriculture on Employment, New Delhi, 1973.
Okai, M., "Some Aspects of Labour Use in the Agriculture of the Lango
District of Uganda," Makerere University College, RDR paper 25,
Orcino, Bert, Economic Aspects of Hand Tractor Ownership and
Operation, Penang, Malaysia, A/D/C Seminar on Farm Mechanization
in Southeast Asia, December 1972.
Orcino, Norberto and B. Duff, Technical and Economic Characteristics
of Tractor Contract Operations in Central Luzo, Los Banos,
Philippines, IRRI, June 1973.
Orcino, Norberto and B.Duff, Expermental Results from Alternative
Systems of Land Preparation, Los Banos, Philippines, IRRI, July
Parthasarathy, G. and.V. Abraham, "Tractorization of Agriculture:
Employment and Productivity -- An Empirical Enguiry in West
Godavari District," (Preliminary Abstract), Department of Cooper-
ation and Applied Economics, Andhra University, Waltair, India 1974.
Pathak, B.S., P.K. Gupta, and B.C. Thakur, Energy Requirements in
Agriculture, Mimeograph No. FM 13-72, Agricultural Engineering
Department, Punjab Agricultural University, Ludhiana, Punjab, India, 1972.
Porter, D.N., Survey of Farmers: Working Paper for Agricultural
Machinery Sector Study, UNDP/IBRD Technical Assistance Project
(Philippines), February 1974.
Porter, D.N. and R.R. Derrick, Some Policy Issues for the Agricultural
Machinery Sector: Working Paper for Agricultural Machinery Sector
Survey, UNDP/IBRD Technical Assistance Project (Philippines),
Rao, C.H. Hanumantha, "Farm Mechanization in a Labour Abundant Economy,"
Economic and Political Weekly, VII: 393-400, February 1972.
Robinson, R.D., "Tractors in the Village -- A Study in Turkey," Journal
of Farm Economics, XXXIV: 451-462, 1952.
Ross, Vernon E., Historical Development of the IRRI Cooperative Rainfed
and Upland Rice Applied Research Project, Los Banos, Philippines,
IRRI, February 1974.
Ross, Vernon E., Asian Rice Farmers Can Harvest Two Crops of Rainfed
Rice by Direct Seeding, Los Banos, Philippines, IRRI, July 1974.
Royal Thai Government, Thailand Farm Mechanization and Farm Machinery
Market, Bangkok, Thailand, 1969.
Sanders, J., The Economics of Agricultural Mechanization in Brazil,
unpublished Ph.D. dissertation, University of Minnesota, 1973.
Sarma, R.K., Economics of Tractor versus Bullock Cultivation,
Agricultural Economics Research Centre, Delhi, India, 1972.
Schmitz, Andrew and David Seckler, "Mechanized Agriculture and Social
Welfare: the Case of the Tomato Harvester," Amer. J. of Agr.
Econ, LII: 569-577, November 1970.
Sethuraman, S.V., "Mechanization, Real Wage and Technological Change
in Indian Agriculture," Manuscript, East-West Center, University
of Hawaii, 1973.
Sidhu, Surjit Singh, Economics of Technical Change in Wheat Production
in the Punjab (India), Ph.D. Thesis, University of Minnesota,
Singh, Gajendra, Energy Inputs and Agricultural Production Under
Various Regimes of Mechanization in Northern India, Ph.D. Thesis,
University of California, Davis, June, 1973.
Singh, Gajendra and W. Chancellor, Relations Between Farm Mechaniz-
ation and Crop Yield for a Farming District in India, Lexington,
Kentucky, Annual Meeting American Society of Agricultural Engineers,
Singh, Gajendra and W.Chancellor, Energy Inputs and Agricultural
Production Under Various Regimes of Mechanization in Northern India,
Stillwater, Oklahoma, Annuall Meeting American Society of Agricultural
Engineers, June 1974.
Southworth, Herman, ed., Farm Mechanization in East.Asia, New York,
The Agricultural Development Council, 1972.
Stevens, Robert D. and Bashir Ahmad, "Agricultural Mechanization and
Rural Welfare in South and Southeast Asia," Agricultural Mechaniz-
ation in South East Asia, 57: 29-32, Spring 1971.
Stout, B.A. and C.M. Downing, Selective Employment of Labor and
Machines for Agricultural Production, East Lansing, Michigan,
Institute of International Agriculture, Monograph No. 3, April 1974.
Thirsk, Wayne R., "Income Distribution, Efficiency and the Experience
or Colombian Farm Mechanization," Paper No. 33, Houston, Texas
Program of Development Studies, Rice University, 1972.
Thirsk, Wayne R., The Economics of Farm Mechanization in Colombian
Development, Ph.D. dissertation, Yale University, 1972.
Timmer, C.P., "Employment Aspects of Marketing Change in Indonesia,"
Food Research Institute Studies, XI: 59-88, 1972.
UNIDO, Report of the Joint UNIDO/IRRI Expert Group Meeting on the
Design and Manufacture of Wet-Land (Rice) Mechanization,
Harvesting and Threshing Machinery in Developing Countries of
Asia and the Far East Region, Los Banos, Philippines, UNIDO,
Van Wersch, Herman J., "Rural Development in Morocco: Operation
Labour," Economic Development and Cultural Change, XVII: 33-49,
Venkatappiah, B., "Issues in Farm Mechanization," Teaching Forum,
No. 11, New York, The Agricultural Development Council, January 1972.
Verma, S.R., "Bird's Eye View of Agricultural Machinery Research
and Development in India," Agricultural Mechanization in Asia,
summer issue 1972.
Voss, C., "Agricultural Mechanization, Production and Employment,"
Monthly Bulletin of Agricultural Economics and Statistics, AO,
Rome: Vol. 23, No. 1, January 1974.
Well, P.M., "The Introduction of the Ox Plow in Central Gambia,"
African Food Production Systems: Cases and Theory, P.M.F.
McLoughlin, Ed., John Hopkins Press, 1970.
Wickham, T.H., The Effects of Irrigation on Rice Yields, Los Banos,
Philippines, IRRI, April 1974.
Wickham, T., 0. Giron, A. Valera, and A. Mejia, A Field Comparison
of Rotational and Continuous Irrigation in the Upper Pampanga
River Project, Los Banos, Philippines, IRRI, August 1974.
Winkelman, D., "The Traditional Farmer: Maximization and Mechanization,"
Development Centre Studies, Employment Series No. 7, Paris, O.E.C.D.,
Yudelman, Montague, Gavan Butler, and Ranadev Banerji, Technological
Change in Agriculture and Employment in Developing Countries,
Paris, Organization for Economic Co-operation and Development,
ARC No. 301.35 M571