ArticlePDF Available

Agricultural Mechanization: A Comparative Historical Perspective

Authors:
Agricultural Mechanization: A Comparative Historical Perspective
Author(s): Hans Binswanger
Source:
The World Bank Research Observer,
Vol. 1, No. 1 (Jan., 1986), pp. 27-56
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/3986307 .
Accessed: 06/08/2014 10:07
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The World
Bank Research Observer.
http://www.jstor.org
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
AGRICULTURAL
MECHANIZATION
A Comparative
Historical
Perspective
Hans Binswanger
he mechanization
of
farming
in
developing
countries has
been
very
uneven. In certain
parts
of
Africa,
in
Java,
and
in
many
hilly regions,
farmers still till their fields
with hand tools
even
though animal tillage has been
common in other
parts
of the world for
thousands
of
years. While draft animals have completely disappeared
in
North
America, Europe,
and
Japan, they
have been
widely accepted
in
Senegal only
in
the
past
few decades. Even in countries
where
farming
is
beginning to be mechanized, power tillers and tractors are still restricted
to
tillage and a few other operations.
This
paper discusses the history of mechanization, the major reasons
for
the wide diversity observed, the options for developing countries in
extending mechanization, and the role of government policy in influenc-
ing
the choice
of technology. The emphasis is on the adoption of me-
chanized
techniques
in
farming systems which are already using animal
draft. The
issues surrounding the introduction of animal draft where
only
hand cultivation
is practiced are discussed in Pingali, Bigot, and
Binswanger (1985). Instead of a summary or conclusions, a set of gener-
alizations is
presented in the text.
Economic
Influences
The
pattern
and
speed of
mechanization is
heavily
influenced
by
rela-
tive
scarcities
of
capital and
labor, and other
macroeconomic
variables.
The
responsiveness of
invention and
innovation to
economywide
factors
has
become
known as
the
process
of
induced
innovation
(Hayami
and
Ruttan
1971;
Binswanger and
Ruttan
1978).
Generalization 1. The
rate
and
pattern
of
mechanization
are
governed
substantially
by
an
economy's land
and
labor
endowments, by
the nona-
gricultural
demand
for
labor, and
by demand
for agricultural
products.
27
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
The
history
of
agricultural growth
and
mechanization
in some
of
today's
industrial countries illustrates
this
generalization.
In
1880,
factor
endowments differed
widely among
these
countries,
with
Japan having
only
0.65 hectares of land
per
male worker
and the United States about
forty
times as much
(see Binswanger
and Ruttan
1978,
tables 3-1
and
3-2).
European
countries
fell in
between,
with
land in the United
King-
dom about twice as abundant as on the
continent. These differences in
endowments
were reflected in massive differences in factor
prices. In
Japan,
a worker had to work
nearly 2,000 days
to
buy
a hectare of
land,
while
his
counterpart
in the
United States
needed to work
only
one-tenth
of
that time.
During
the
ninety years
until
1970,
land-labor ratios
increased in all
countries, especially
after 1950. These increases
reflect
the
rapid
declines
in
agricultural
labor
forces as
people
moved to
industry
and off the
land.
The
United States, however,
had an
increase in
agricultural
land as well
as a reduction
in
the
agricultural
labor
force,
so that differences
in
land-labor ratios between it
and
other
countries increased.
Despite
their
differences
in
natural
resources, Japan,
European countries,
and
the
United States
managed
to
expand
their
agricultural
output by up
to 1.7
percent
a
year. Japan
and
the
continental
European
countries
achieved
their
rapid growth
because
yields (output
per
hectare of
arable
land)
grew at about 1.5
percent
a
year,
or
roughly
twice as
fast
as in
the
United States.
Japan
and the United
States
relied on different
technological paths
to
expand
their
agricultural output.
Research
summarized
by Hayami and
others
(1975)
and
Binswanger
and Ruttan
(1978)
has
established that
Japan
has
long emphasized biological,
yield-raising technology,
much of
it
supported by heavy
investment in
irrigation.
This
emphasis
continued
with
systematic
investment in
agricultural research
initiated
after
1868.
Until the 1950s mechanization
played only
a minor role
(see
table
1).
The
emphasis
on
biological technology
was
supported by
conscious
government
choice: in the
late nineteenth
century Japan imported machi-
nery
from
the United States, but did not find
it
useful. It
then hired
biologists
from
Germany
to
assist
in
developing
its
biological research
program, which was successful. The United
States, however, emphasized
mechanical
technology
even
before 1880 (see
tables 2 and 3). Although
publicly
funded
biological research was
initiated in the 1870s, it did not
produce big increases in yields until about
1930, well after the major
land
frontiers had
been closed and
mechanization was far advanced.
Successful
agricultural growth
in
various
developed countries has
therefore
capitalized on abundant factors of
production: land and me-
chanization in the
United States; labor, land
improvements, and biologi-
cal
technology
in
Japan. Continental Europe
also emphasized biological
technology
before
shifting the emphasis to
mechanical technology.1
28
Research
Observer 1, no. 1 (January 1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
z
Table
1.
Pattern
of
Agricultural
Mechanization
inJapan
(thousands)
Draft
Number
and
Power
of
beef
Rice
sprayers,
Power
Riding
Rice
Year
farms
cattle
Horses
Motors
Pumps
Thresbers
hullers
dusters
Cultivators
tillers
tractors
Binders Combines
transplanters
1880
5,500
1,152
1,626
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
1900
5,502
1,204
1,542
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a. n.a.
n.a. n.a.
n.a.
1910
5,518
1,259
1,564
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
1920
5,564
1,256
1,468
2
2
0.5
0.6
n.a.
n.a.
n.a. n.a.
n.a. n.a.
n.a.
1931
5,632
1,361
1,477
92
28
56
77
0.2
n.a.
n.a.
n.a.
n.a. n.a.
n.a.
1939
5,492
1,767
1,168
293
83
211
132
5
3
3 n.a.
n.a. n.a.
n.a.
1945
5,670
1,827
1,049
424
87
364
177
7
8
7
n.a.
n.a.
n.a.
n.a.
1951
6,145
_a
1,112
1,295
92
1,080
460
20
29
16
n.a. n.a.
n.a.
n.a.
1955
6,027
n.a.
888
2,140 122
2,060
700
87 82
82
n.a.
n.a. n.a.
n.a.
1960
5,966
n.a.
618
2,799
288
2,651
878
305
791
514
n.a.
n.a. n.a.
n.a.
1966
5,665b
n.a.
396c
3,108b
n.a.
3,172
1,008b
1,126
n.a.
2,725
39
146b
n.a.
n.a.
1971
5,342b
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2,400
n.a.
3,201
267
582 84
46
1976
4,835b
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2,898
n.a.
3,183 721
1,498 428
1,046
1979
4,742
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
2,618
n.a.
3,168
1,096
1,704
747
1,601
n.a.
Not
available.
a.
Continued as beef cattle.
b.
Figure
corresponds
to nearest
adjacent year.
c.
Figure
corresponds to 1963.
Sources:
Kazushi
Okawa,
M.
Shinohara,
and
M.
Umemura, Estimates
of
Long-Term Economic
Statistics
of
Japan
since
1868:
Agriculture
and
Forestry,
no.
9
(Tokyo,
1966);
and Farm
Machinery
Statistics
(1981).
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table
2.
Sources of
Farm
Power
in
the
United
States
(thousands)
Tractors
(exclusive of
steam and
garden)
Number
Workstock above
Horse-
of
two
years
Steam Gas
power
Year farms
Oxen Mules Horses Windmills
engines
engines
Number
(millions)
Trucks
1870
2,660
1,319 1,125 7,145
n.a. n.a. n.a.
n.a.
n.a.
n.a.
1880
4,009
994
1,813 10,357
200
24
n.a.
n.a.
n.a.
n.a.
1890
4,565 1,117
2,252
15,266
400 40
n.a. n.a. n.a.
n.a.
1900
5,737
960
2,753
15,506
600
70 200
n.a.
n.a.
n.a.
1910 6,406 640
3,787
17,430
900 72
600 10 0.5 0
1920
6,518
370
4,652
17,221
1,000
70
1,000
246
5
139
1930
6,546
n.a.
17,612a
n.a.
1,000
25
1,131 920
22
900
1940
6,350
n.a.
13,029
n.a.
n.a.
n.a. n.a.
1,567
62b
1,047
1945
5,967
n.a.
11,116
n.a.
n.a. n.a.
n.a.
2,354
88c
1,490
1950
5,648
n.a.
7,415
n.a.
n.a. n.a.
n.a.
3,394
93
2,207
1955
4,654
n.a.
4,101
n.a. n.a.
n.a. n.a.
4,345
126
2,675
1959
4,105
n.a. n.a.
n.a. n.a. n.a. n.a. n.a.
n.a.
n.a.
1960
3,963d
n.a.
2,883
n.a. n.a. n.a.
n.a.
4,685
153
2,826
1965
3,356
n.a.
_e
n.a.
n.a.
n.a. n.a. 4,787
176
3,030
1970
2,949
n.a.
n.a.
n.a. n.a.
n.a.
n.a.
4,619
203
2,984
1975
2,767
n.a. n.a. n.a. n.a.
n.a. n.a.
4,469
222
3,031
1979
2,672f
n.a. n.a. n.a. n.a. n.a.
n.a.
4,3509
243
3,045
n.a. Not available.
a.
From 1930 onward refers to total workstock on
farm.
b.
Average
horsepower
for 1930-34
multiplied
by
number of tractors
in
1930.
c.
Average
horsepower
for 1940-44
multiplied
by
number
of
tractors
in
1940.
d. After 1960
corresponds
to
1969 definition.
e.
Discontinued.
f.
Figure
corresponds
to 1978.
g.
Tractors
over
40
horsepower only.
Sources:
Number of farms:
up
to
1959,
U.S.
Department
of
Agriculture,
Century of Agriculture
in
Charts
and
Tables;
1960-79: U.S.
Department
of
Commerce,
Statistical Abstract
of
the United
States
(1980).
Oxen,
mules, horses,
windmills, gas
engines,
and
steam
engines,
1850-1930:
W.
M.
Hurst and L.
M.
Church,
Power and
Machinery
in
Agriculture
(1933),
table
8, p. 12;
1930-79:
U.S.
Department
of
Commerce,
Historical Statistics
of
the
United
States: Colonial Times to 1970
(1975).
Tractors,
horsepower,
and
trucks,
1870-30:
W. M.
Hurst and L. M.
Church,
Power and
Machinery
in
Agriculture
(1933),
table
8, p.
12;
1940-59: U.S.
Department
of
Agriculture, Changes
in Farm
Production
and
Efficiency,
1964
and
1973;
1960-79:
U.S.
Department
of
Commerce,
Statistical Abstract
of
the United States
(1980).
Generalization
2.
Mechanization
leads
directly
to increased
yields
only
in
exceptional
circumstances,
such as
when
high-yielding seeds,
pesti-
cides,
and
fertilizers
are also used.2
Thus,
extra
machinery usually
sub-
stitutes
for
labor
or-where
they are
already in
use-for
animals.
This
generalization
corresponds to
the
substitution
view of
agricultur-
al
mechanization
(Binswanger 1978). It
differs
from the
net
contributor
view,
which
assumes
that
more
machinery-in
particular,
tractors-
30
Research
Observer 1,
no.
1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 3.
Production or Sales
of Horse-Drawn and
Tractor-Drawn Machines
in the United States
(thousands)
Corn
surface
planters
only
Number Harrows
(hand
Self
Threshers Hay
making
of
Plows (all Seed drills and rake Grain
Horse
Steam
Horse
Year?
farms Horse
Tractor
types)
Cultivators Horse
Tractor horse)
reapers
binders
(small) (large)
Combines
Mowers rakes Loaders Stackers
1870
2,660
865 n.a. 9
89 n.a. n.a. 22
60 n.a. 23
n.a. n.a.
81 n.a.
n.a.
1880
4,009 1,326
n.a. 128
318 n.a. n.a. 69
35 n.a. 10
n.a. n.a. 96
9
n.a.
1890
4,565
1,249
n.a. 269
445 n.a. n.a. 132
9 n.a. 11
n.a. n.a. 115
3 n.a.
1899b
5,737c
1,075 n.a. 478
505 n.a. n.a. 208
36 n.a. 1.3 3.6 n.a.
n.a.
216
7 n.a.
1899d n.a.
973 n.a. 478
296
92
n.a. 208
36 233 1.3 3.7
n.a. 399
216
7
12
1909
6,406e
1,358 n.a. 701
435 68 n.a. 219
58 129 2.2 8.0
0.5
359
266 35
17
1920
6,518
714 145 604
579 107 3 132
2 100 16.5 4.2
2.7 173 118
32 10
1929
6,512
324
117 540
398 36 16 93
n.a. 65 9.6 1.3
19.6 115
91
26
6
1938
6,527
137 124 351
214 28 57
n.a. 31 2.7 3.6
41.5 76
54 19
1
n.a.
Not
available.
a.
Figures
for years before 1920 represent numbers
manufactured. The earliest sales
figures available are for 1920.
b.
Data from U.S. Department of
Commerce.
c.
Figure
corresponds
to 1900.
d. Data
from McKibben, Hopkins,
and
Austin.
e.
Figure
corresponds
to 1910.
Sources:
1870-99: U.S.
Department
of Commerce, Historical Statistics of the
United States: Colonial Times to 1970.
1899-1938: E. G. McKibben,J.
A.
Hopkins,
and
Griffin
R. Austin, Changes
in Farm
Power
and
Equipment Field Implements (Philadelphia:
Work Projects
Administration, National Research Project,
August 1939).
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
produces
higher
yields
or
other
gains
in
output,
regardless
of the
econ-
omic
environment
in
which
it is
introduced.
Such a view
usually
confus-
es
the
direct
effects
of
mechanization
with
the indirect
productivity
effects of factor
savings.
For
example,
in an
extensively
farmed
area
of
Africa where hoe cultivation
is used
yields
may
be
low,
while
in
an
intensively
farmed
tractorized
region
of India
yields
may
be much
higher.
The
yield
differences
may
be caused
in
part
by
differences
in
other
inputs,
such as fertilizers
or seeds.
They
could
also be caused
by
better
tillage in India-but
this does not
mean that
good
tillage
is
achieved
only
by
tractors and
cannot
be achieved
by
hand.
Examples
from
Java
show
that cultivation
by
hand
can be as
thorough
as
by
ox or
tractor.
The lower
tillage
intensity
in Africa
may
simply
reflect
the abundance
of
land:
in order to maximize labor
productivity,
people
work
thinly
over
a
large area.
Under the substitution
view,
the
profitability
of
mechanization
and
its
contribution to economic
growth
depend on the
opportunities
available
to workers
(and
sometimes draft
animals)
released from
their
tasks.
Hence the
third
generalization:
Generalization 3.
Mechanization
is most
profitable
and
contributes
most to
growth
where land is
abundant,
where
labor is
scarce
relative
to
land,
and
where labor
is
moving
rapidly
off
the
land.
Several
cases,
listed in table
4,
illustrate the
effects of
mechanization
on
employment.
In
case
1
unused
land is available
and
mechanization
leads to
output
growth-the
more
so,
the
higher
the
elasticity
of
final
demand.3 The
best
example
is the
United States in
the
second half
of
the
century:
an
impressive
horse-based mechanization
led to
massive
agricul-
tural
growth
because land
was
rapidly
opened
up
and
export
markets
in
Europe provided a
highly
elastic demand for
agricultural
products.
Total
farmland
more
than doubled
between
1870 and
1920.
Since
average
farm
size
stayed
roughly
constant,
total
farm
employment
must
have
nearly
doubled as well.
The
agricultural work
force,
far from
being
displaced,
was
redeployed within
agriculture,
along
with
large
numbers
of
immi-
grants.
Mechanization
did not
produce
higher
yields,
however;
they
came later
(see table
5)
and
were
linked to
fertilizers
and
biological
innovations.
In
all
these
changes, the
elastic
demand
provided
by
export
markets
played
a
crucial role.
Without
such
export
possibilities,
areas
planted,
employment, and
agricultural
output
would
have
expanded
less
and
mechanization
would
probably
have
happened
more
slowly.
(If
final
demand
is
very
inelastic,
mechanization
could
lead to a
reduction
in
agricultural
employment even
if
extra
land is
available.)
Mechanization
can
also
be
induced
by
labor
scarcity
arising
out
of
nonagricultural
demand for
labor
(case
2).
Production
costs rise
because
32
Research
Observer
1, no.
1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 4. Direct
and
Indirect
Effects of
Agricultural
Mechanization
Indirect
Indirect
Forces
Immediate
effect
effect
leading
consequence
on
on
to
of
agricultural
agricultural
mechanization
mechanization
output
employment
Examples
1.
Land available
Labor used
on
larger Expands,
and more
Expands
if
demand
is
Nineteenth-century
areas, production
quickly
the more
elastic; stagnates
or United
States
costs drop
elastic is
final falls
if
demand
is
demand
inelastic
2. Wages rising in Production costs
rise Falls
(or
grows
more
Falls
United States after
response to
less
than in
absence slowly),
but
by
less
1940; Japan, Europe
nonagricultural
of
mechanization
than
in
absence
of
after
1955
labor demand mechanization
3. Unmechanized
A
new method of
Expands,
and more
Expands,
and more
Pumping
in
technique production
becomes quickly
the more
quickly
the
more
contemporary
Asia
unprofitable
profitable elastic is
final
elastic is
final
demand
demand
4.
Subsidies on
capital
Production costs
may
Small
expansion
at
Falls,
sometimes
Contemporary
energy
drop
modestly
or
best
sharply
Brazil, Pakistan,
stay constant
China
wages
rise
rapidly.
Other
things being equal,
farming output
will there-
fore
fall
(or
grow more
slowly), depending
on the
elasticity of final
demand.
Farmers
mechanize,
although they
can seldom
prevent
some
increase in
their production
cost.
The
best
example of these trends
comes from the
United
States
after 1940. The
use
of
tractors,
combines,
and other
machines
expanded
at
unprecedented rates
(see
table
2).
Al-
though labor
input per
acre or
per
animal
had declined
a
little
between
1915 and
1939,
it
fell
sharply after 1940
(see table 5).
Agricultural
employment also
fell
substantially, both
in
absolute and
relative terms,
and
labor was
redeployed
outside agriculture.
The number of
workers
per
farm
was
stable,
while
farm sizes grew
rapidly from an
average of
167 acres in
1950 to 401
acres in 1978.
Europe went through
equally
dramatic
changes
after
1955.
Cases 1 and 2
show
that the
labor
effects of
mechanization
depend on
the
alternatives available to the
economy. The
Indian Punjab
provides an
opposite
example. The green
revolution
initiated in the
mid-1960s led to
sharply
increased demand for
labor, which
caused a big rise
in real
wages
around
1968
(Gupta
and Shangari
1979). This in
turn led to
increased
seasonal and
permanent migration,
primarily from
Eastern
India.4 But it
also led to the
use of more
tractors and
threshers by
Punjab
farmers. The combined
effect of these
developments
was a dec-
line in
real
wages after
1972-73, which
brought them
closer to the
stagnant
real
wages in the
rest of India's
agriculture.
Because India's
Hans
Binswanger
33
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 5. Productivity
Indicators: Labor Hours
per
Unit of Production and
Related Factors,
Selected
Crops
and Livestock in the United
States,
1915-78
Crop and livestock 191S-19 1925-29
1935-39 1945-49 1955-59 1965-69
1974-78a
Crop
Corn for
grain
Hours
per
acre
34.2 30.3 28.1 19.2 9.9 5.8
3.7
Yield
(bushels)
25.9 26.3 26.1 36.1 48.7
78.5
87.8
Sorghum grain
Hours per
acre
n.a. 17.5 13.1 8.8 5.9 4.2
3.9
Yield
(bushels)
n.a. 16.8 12.8 17.8 29.2 52.9
50.8
Wheat
Hours per
acre
13.6
10.5
8.8
5.7
3.8 2.9
2.9
Yield
(bushels)
13.9
14.1 13.2 16.9 22.3 27.5
30.0
Hay
Hours per
acre
13.0 12.0 11.3
8.4 6.0
3.8
3.5
Yield
(tons)
1.25 1.22 1.24 1.35 1.61 1.97 2.15
Potatoes
Hours per acre
73.8 73.1 69.7 68.5 53.1 45.1 38.3
Yield
(cwt.)
56.9
68.4 70.3 117.8 178.1 212.8
257.0
Sugarbeets
Hours per acre
125
109
98 85
51
33
26
Yield (tons)
9.6 10.9
11.6
13.6 17.4
17.5 19.7
Cotton
Hours per
acre
105
96 99 83
66
30
10
Yield
(pounds)
168 171
226 273 428 484
462
Tobacco
Hours per acreb 353 370 415 460 475 427 259
Yield
(pounds) 803
772 886
1,176 1,541 1,960 2,049
Soybeans
Hours per
acre 19.9
15.9 11.8 8.0 5.2 4.8 3.7
Yield (bushels) 13.9 12.6 18.5 19.6 22.7 25.8 27.8
(continued)
economy
was
growing slowly,
a slower rate of mechanization
and a
larger
volume
of
migration
could have solved
the labor
shortages
in
Punjab
at a lower
capital
cost. And
the
extra
employment would have
meant that the benefits of
the green
revolution were shared more
widely
with
workers
in
poorer regions.
Mechanization can be a
powerful
stimulus
to
growth
when it
makes a
new
method
or
crop profitable (case 3). The best example is pump
irrigation. Although
it
is
always possible
to lift
water with
animal or
human
power,
it
may
often not be
profitable to do so, even at extremely
low
wages.
The
pump
therefore enables
output to rise but the size of
the
increase
is
determined
by the elasticity of final demand. Since the
extra
production requires
extra
labor, agricultural employment expands
more or less in
step with output.
34 Research Observer
1,
no.
1
(January 1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table
5
(continued)
Crop and livestock
1915-19 1925-29 1935-39 1945-49
1955-59 1965-69
1974-78a
Livestock
Milk cows
Hours per
cow
141 145 148
129 109 78
48
Milk
per cow (pounds)
3,790
4,437
4,401
4,992
6,307 8,820 10,783
Cattle other
than
milk
cows
Hours per
cwt. of beef
producedc'd
4.5
4.3
4.2 4.0 3.2 2.1
1.4
Hogs
Hours per
cwt.
producedd
3.6
3.3 3.2 3.0
2.4 1.4
0.6
Chickens (laying
flocks and
eggs)
Hours per 100 layers n.a.
218
221
240 175
97
61
Rate of lay
n.a.
117
129 161 200
219
234
Chickens (farm raised)
Hours per 100 birds
33
32
30
29 23
14
12
Hours per
cwt.
producedd
9.4
9.4
9.0
7.7 6.7
3.7
3.0
Chickens
(broilers)
Hours per 100 birds
n.a. n.a.
25 16
4 2
0.6
Hours per cwt. producedd
n.a.
n.a.
8.5
5.1 1.3 0.5 0.2
Turkeys
Hours per cwt. producedd
31.1 28.5
23.7
13.1
4.4 1.3
0.6
n.a.
Not available.
cwt.
Hundred
weight.
Note:
Labor hours per
acre
harvested include preharvest
work on
areas
abandoned, grazed,
and turned under.
a.
Preliminary.
b.
Per acre planted and
harvested.
c.
Production includes beef
produced as
a
by-product of the
milk
cow enterprise.
d.
Liveweight production.
Source:
Economics,
Statistics and
Cooperative Service-Economics.
The
pace
of
mechanization is influenced
by
three
other economic
factors:
capital scarcity
and
energy costs,
farm
size,
and subsidies.
Capital scarcity
and
energy
costs. Poor societies have
smaller
capital
stocks than rich
ones, and
the cost of capital (in terms
of labor) is
higher. High capital costs retard mechanization in
several
ways. First,
they reduce the
profitability of all forms of agricultural
investment,
including land
improvements, irrigation, animals, and
buildings. Second,
they may
cause
farmers
to allocate whatever investment funds
are avail-
able
away
from
mechanical inputs. This trend will be
stronger the more
expensive
and
long-lived the mechanical inputs are and the
easier it is to
produce
other
forms of
capital (such as land improvements)
by hand. A
third
effect, discussed
in
detail in the next section, is that
higher capital
costs
produce
a bias in
mechanization toward
power-intensive opera-
Hans
Binswanger
35
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
tions.
Finally, higher
capital
costs influence
the
design
of
machines;
if
repair
costs
are
relatively low, designs
emphasize
repair
over
durability.
Generalization 4.
High capital
costs
(relative
to
labor)
retard
mechani-
zation
and
lead to
selective
emphasis
on
power-intensive
operations.
Machinery
design adjusts
to
high
capital
costs
by
lack
of
convenience
features,
simplicity, and reduced
durability.
Energy
is
only
one
of
the
costs
of
using
machines.
Capital
and
mainte-
nance costs
are
often
larger.
Since
the
profitability
of
machines-their
comparative
advantage-is
tied
closely
to
labor
costs, expensive
energy
is likely
to retard mechanization
much more in
countries with
cheap
labor.
Farm size.
The
size
of the
average
farm is
largely
a
reflection of
the
scarcity
of land relative to
labor
and thus
need
not
be
an
independent
influence
on the
pace
of mechanization. Mechanization
can
certainly
facilitate
the
growth
of
large farms,
however,
as it did in
the United
States
after
1940 and
later
in
Western
Europe.
Generalization 5.
Mechanization
is the
main
facilitator
of
the
trend
toward
bigger farms.
This
goes
along
with another
lesson
from
historical data and
contem-
porary experience:
Generalization 6.
Large farms
adopt
new
forms
of
machinery
consi-
derably faster
than small
farms.
Because
larger
farms offer more
collateral, they
make it
easier to
bor-
row to invest
in
new
machinery.5
In
addition,
some
(but
not
all)
mechani-
zation is
subject
to
genuine
economies
of
scale:
it
is
technically
more
efficient to
design
a
large
rather
than a small machine.
Even
machines
invented
in countries
with abundant labor
(and
therefore
smaller
farms)
were first
developed
for the
largest
farms,
because
they
had
the
lowest
costs
of
capital
relative to labor.6
The market for
machines
expanded to
smaller
farms
only
when labor
costs rose or
capital
became
more
abun-
dant. In the
history
of
engineering,
technical
developments
have
often
been embodied in
smaller
and smaller
machines.7
Japan,
in
particular, has
developed
many
machines for
small
farms
and
plots.
For
certain
operations,
mechanization
spreads to
small
farms when
machinery
can
be rented
rather than
bought.
For
a
rental
market to be
established,
the
optimal
farm size for
owning
a
machine
must
be bigger
than that of
numerous small
farms. In
addition, it
is
easier to
establish
rental markets for
operations that do
not need to
be done on
all farms at
the same
time:
threshing and
milling
are
examples. It is thus
no accident
36
Research Observer
1,
no.
1
(January 1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 6. Ownership
and Use of
Farm
Equipment
in
the
Philippines,
1971
(thousands)
Number
Number
of
machines Number
offarms
using
machines Ratio
Type of offarms
owned
by Ownedfully
Rented or
provided of
renters
equipment reporting
farm
operator
or
party by
landlord to
owners
Total number of farms 2,355
-
-
-
-
Plows 1,170 1,511 1,366
129 0.09
Harrows
887
1,069 1,031
94
0.09
Tractors
11
16
12
78
6.50
Stripping machines,
crushers, shellers
16 19
18 85
4.72
Harvesters and threshers
14
26 16 132
8.25
Power-producing machines
5 7 6 3
2.00
Carts and
wheelbarrows 262
292 306
46
0.15
Motor vehicles
14
19 15 69 4.60
Sprayers 79
90
89 61 0.69
Source:
National Census and Statistics Office, Philippine Census of Agriculture 1971 (Manila: National Economic and
Development Authority, 1971).
that rental
markets
for
threshing
machines were
well established
in
the
nineteenth
century
in the United States
and are now common
all over
Asia (Gardezi and others 1979;
Walker and
Kshirsagar 1981).
The
contract-hire system for combines
in the United States
illustrates
the
problem
of
synchronized timing. The contractors
achieve
higher
rates
of
machinery
utilization
by migrating
to
follow
the
harvest from the Texas-
Oklahoma area to
the
northern
states,
where
harvesting
takes
place
months
later.
Milling
rice for home
consumption
can
also
be done over a
long period,
and in
Asia
it is
common for a mill owner
to "rent" his
machine to
customers. Rental markets
for land
preparation,
with the use
of
animals
or
tractors,
were
once common
in
the United
States,
when
plows
were
scarce,
and in
Europe.
Such rental markets are
now common
in Asia wherever
tractors
or
power tillers
have
penetrated.
These
characteristics of rental markets are confirmed by some
data for
tractor rentals in South Asia and
machinery
rentals in the
Philippines
(see
table
6).
The
Philippine
data show
that
most farmers own their
animals, carts, plows, and harrows. Harvesting and threshing equip-
ment, tractors,
and motor
vehicles, however,
are used on about five to
seven
times more
farms than own them, which indicates that rental
markets are
extremely well developed.
Generalization
7.
Where
rental
markets are fairly easy to establish,
farm size has much less influence on the size of machines.
Subsidies.
Subsidies may speed up mechanization (case 4). Because the
direct
effects of
mechanization
on
yields are small, however, any effect of
Hans
Binswanger 37
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
subsidies
on
agricultural output
must be an
indirect one that
arises from
the cost reduction
made
possible
by
machines.
But when mechanization
is not spontaneously
driven
by
some
form
of labor
scarcity,
the
impact
on
production
costs is
not
large;
the
output
effects of subsidies therefore
cannot
be
large
either.
When mechanization is caused
by subsidies,
reductions
in
the agricul-
tural work force can
be
substantial.
Unlike
in
cases
1
and
2,
workers
who lose
their
jobs
will
find
only
inferior
alternatives,
and some
may
remain
unemployed;
this redeployment
of labor is not a
productive
benefit,
but
a
loss. Nor
is
there
any potential
relief from
drudgery
for
the
redeployed workers,
since their inferior
work
options may
in
fact
entail
more
drudgery.
Patterns
of
Mechanization
The
most dramatic
aspect
of mechanization
is the shift
from
one
source of
power
to another. In ancient China cattle
began
to
replace
human labor more
than
3,200
years
ago.
Between the
second and
fourth
century
A.D.
fairly
widespread
use
of
water
power
is
reported
from
China
for
rice
pounding,
grinding,
and
water
lifting (Liu
1962).
Water
power
was
widely
used for
milling
purposes
in
Europe
during
the
Mid-
dle
Ages;
at about the
same time wind
power
is
reported
to
have
been
used in
China and
Europe
as well.
In
nineteenth-century
Europe
and
North
America,
oxen were
displaced
by
horses,
which
provided
power
for
many
mechanical
devices
from
about 1850 to
as late
as 1965.
Steam
engines
were
widely
used for
only
about
fifty years
between
1870 and
1920. After
1900
they
were
rapidly
displaced
by
internal
com-
bustion
engines
and electric
motors. Tractors
came into
widespread
use
in North
America
after
about
1920,
but
coexisted with
horses for
rough-
ly
twenty-five
to
thirty years.
Except
for Great
Britain,
where
tractors
began
to
be
adopted
in
the
1930s,
the
tractorization of
European
and
Japanese
agriculture
was
delayed
until
about
1955,
after
which
it
hap-
pened
very
quickly (table
7).
The
emphasis
on
shifts
in
power
sources,
especially
the
shift
to
trac-
tors,
can
cause
misunderstandings about which
operations are
the
most
likely candidates for
mechanization
in
developing
countries.
This
section
therefore
discusses
mechanization in
terms of
operations
and
pays
only
occasional
attention
to
power
sources.
Most of the
evidence
comes
from
machinery
stock
data;
though
lacking
detail, no other
data
can
give
so
comprehensive a
picture
over
long
periods of
time.8
Operations
can be
grouped
in
terms of
the
intensity
with
which
they
require
power
(or
energy)
relative
to
the
control
functions of
the
human
mind
(or
judgment).
Regardless
of
the
stage
of
mechanization,
new
power
sources are
always
used
first for
power-intensive
operations.
Furthermore, it
appears
that
the
price of
labor
matters
less
for
the
mechanization
of
power-intensive
operations
than
for
control-intensive
38
Research
Observer
1, no.
1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table
7.
Growth
in
Number
of Tractors
in
Selected
Countries
(thousands)
Japan
United
United
Korea
Year
Two wheel
Four
wheel
Germany
Denmark
France
Kingdoma
States
Spain
Yugoslavia
Two
wheel Four wheel
India
Mexico Philippines
1920 n.a.
n.a.
n.a.
n.a.
n.a.
10
246
n.a. n.a. n.a.
n.a.
n.a.
n.a.
n.a.
1930
n.a.
n.a.
n.a.
n.a.
27
30
920
n.a.
n.a. n.a. n.a.
n.a.
4
n.a.
1938-39 3
n.a.
30
4
36
55
1,545
3 n.a.
n.a.
n.a.
n.a. 5
0.2
1945-47
8
n.a.
69
4
77
244 2,613 5 n.a.
n.a.
n.a.
5
n.a.
1
1950 16
n.a.
140
17
137
325
3,394
10 6 n.a.
n.a.
9
23
n.a.
1955 82
n.a.
462
58
305
436
4,345 25 10
n.a.
n.a.
21
n.a.
n.a.
1960
514
n.a.
857
111
680
456
4,688
39
36
1 n.a.
31 55
8
1965-66 2,725
39
1,164
161
996
482
4,787
148 45
11 n.a.
54
n.a.
n.a.
1970
3,201
267
1,371
175
1,230
514
4,619
260
80
44
0
148
91
11
1975-76
3,183
721
1,425
185
1,363
541
4,469
379
226
60
1
228
102
n.a.
1979
3,168
1,096
1,456
190
1,430
508
4,350
492
385
n.a.
n.a.
310
114
n.a.
n.a. Not available.
a. Great Britain
and
Northern
Ireland.
Sources:
Binswanger
(1984),
tables
3, 9, 12,
13, 14,
16, 18;
Organisation
for Economic
Co-operation
and
Development,
Development
of Farm Motorization
and
Consumption
and
Prices of
Motor
Fuels
in
Member Countries
(Paris,
1962); and
Food
and Agriculture
Organization,
Production
Yearbook,
various
issues.
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 8. Machinery
Patterns in France
(thousands)
Hay
and
Sprayers
Cream Root straw
Pick-up Sowing Fertilizer of
Year
Plows
separators
Threshers cutters presses trucks machines distributors
traction
1852 2,578
n.a.
60
n.a.
n.a. n.a. n.a. n.a.
n.a.
1862 3,206a
n.a.
101 28 n.a. n.a. 11 n.a.
n.a.
1882 3,267
n.a. 211 n.a. n.a. n.a. 29 n.a. n.a.
1892 3,669
n.a. 234 n.a. n.a. n.a.
52
n.a. n.a.
1929 1,190b
666
204
n.a.
10 74 322 119 142
1937 n.a.
n.a.
152 n.a.
9 n.a. n.a.
n.a. n.a.
1941
n.a.
n.a.
141
n.a. 9
n.a. n.a. n.a. n.a.
1946
1,325
626
206
1,007
12 n.a. 385 151 85
1950 1,385
686
218
1,099
17 n.a. 410
165 104
1955 1,427
696 215
1,152
26 n.a.
447
221 122
1960
n.a.
672
191
1,152
33 n.a.
514 321
153
1965
n.a.
n.a.
122
n.a.
n.a. n.a.
n.a. n.a. 223
1970
n.a.
n.a.
n.a. n.a.
n.a.
n.a.
n.a. n.a. 304
1977
n.a. n.a.
n.a. n.a. n.a.
n.a. n.a.
n.a. 406
n.a.
Not available.
a. Includes
794
improved plows.
b. Double-sided
plows only
after 1929.
c. Includes
motor-driven mowers.
d.
Only self-propelled
combines.
ones-that
is,
it often
pays
to move to a
higher stage
of
mechanization
in
power-intensive
operations, even
at low
wages, when mechanization
of
control-intensive
operations
is not
profitable.
The rest of this
section
provides
support for the following:
Generalization
8. When
new
power
sources
become available, they are
initially
used
only for selected operations for which their comparative
advantage
is
greatest. Power-intensive operations are shifted most rapid-
ly
to new
power
sources.
Control-intensive operations are shifted to the
more
mechanized techniques
when
wages
are
high
or
rapidly rising.
Power-intensive
processing and pumping. Milling, threshing, chop-
ping, sugarcane
crushing, pumping of water, and the like are extremely
power-intensive but
need little control. Moreover, both stationary and
mobile
sources
of
power can be used for them. Among the stationary
sources,
water
was first
used for milling, pounding, and grinding in the
first
century
B.C.
in
China. Water-powered milling was also invented in
France in
the
fourth
century A.D., though not until the twelfth century
was
it
adopted
throughout Europe. Wind power has historically been
used
almost
exclusively for milling and for lifting small amounts of
40
Research
Observer
1, no. 1 (January 1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Sugar
Reaper!
Motor
Hay Potato
beet
Milking
binders Mowers
mowers Combines
balers
diggers
diggers
Rakes Tedders
machines
n.a.
n.a.
n.a.
n.a.
n.a. n.a.
n.a.
n.a.
n.a.
n.a.
9
9
n.a. n.a.
n.a. n.a. n.a. n.a.
6
n.a.
n.a.
19
n.a. n.a.
n.a. n.a. n.a.
n.a.
27
n.a.
23 319
n.a. n.a.
n.a. n.a.
n.a. n.a.
51
n.a.
420 1,389 n.a. n.a.
n.a.
60 13 739 354
4
341 n.a. n.a.
0.3 n.a.
n.a. n.a.
n.a. n.a.
n.a.
481
1,279
n.a.
0.3 n.a.
n.a. n.a. 733 448
n.a.
501
1,373c
n.a. n.a.
n.a. 67 10
740 n.a.
n.a.
529 1,470c n.a.
5
n.a.
77
11
785
n.a.
46f
560 1,547c n.a.
18 17
90 12 839
n.a.
80
534
n.a.
97
50
51
91
iSe
n.a. n.a.
124
361 n.a. 104 102
169 100 20 n.a. n.a.
186
133
n.a.
105 133
292
92
25 n.a. n.a.
283
n.a.
n.a.
84
148d
445 n.a. n.a. n.a. n.a.
392
e. Includes
only diggers
up
to 1960 and
only complete
harvesters
from 1965 on.
f.
Figure corresponds to
1951.
g. Reaper
binders
only
after
1937.
Source:
Ministere de
l'Agriculture, Statistique
Agricole
(Retrospectifs
1930-1957) (Paris, 1959);
and
Statistique
Agricole
de La
France and
Statistique
Agricole Annuelle (Paris:
Imprimerie Nationale),
various
issues
of
each.
water. Mills
and threshers were the
most common
users
of
steam
power
in
the late
nineteenth
and early
twentieth
century
in both
Europe
and
the
United
States.
Mechanical
threshing based on human
power, but
especially on hors-
es,
became
widespread
in
the United
States and Britain as
early as 1830.
By
1850
virtually all
grain in the United
States was
threshed by large
mechanical
threshers,
which went from
farm to farm during
the winter
months.
Rental
markets were
extensive.
By 1852 the number of
threshing
machines
in
France
had already reached
nearly one-third of its
peak 1929
level
(see table 8),
though they spread
more slowly in
Germany. Except
for some
animal-drawn
primary tillage,
stationary machines
for power-
intensive
operations
preceded all other
forms of mechanization
in Japan.
In
South Asia
animals have long
driven Persian wheels,
sugarcane
crushers,
and oil
crushers. Animals used
in these operations
are increas-
ingly being replaced
by diesel and electric
engines. In India in
1972 the
number
of
stationary
engines for
power-intensive operations
was about
twenty
times
that of
tractors (see table
9). And in China
(table 10) the
number
of
threshers
alone exceeded the
combined total of
tractors and
power
tillers, even in
1980. In all of
Asia mechanical rice
milling for
large
trade
quantities
had already been
introduced in the late
nineteenth
Hans
Binswanger
41
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table 9.
Pattern
of
Farm Mechanization
in
India
(thousands)
Year
1945 19S1
1956
1961
1966
1972
Draft animals
59,333
67,383 70,690 77,986
78,517
80,137
Persian wheel
n.a.
n.a.
n.a. 600
680
638
Oil Pumps
12 83
123
230 471
1,558
Electric pumps
9
26
47 160
415
1,618
Tractors
5
9
21
31
54 148
Plows
Bullock
Wooden
27,306
31,796
36,142
38,372
39,880
39,294
Iron
487
931
1,376
2,298
3,521
5,359
Tractor
n.a.
n.a.
n.a. n.a.
n.a.
57
Other tillage
implements
Bullock
n.a.
n.a.
n.a. n.a.
2,724
17,119
Tractor
n.a.
n.a.
n.a.
n.a.
n.a.
111
Sugarcane
crushers
Power
9 21
23 33
45
87
Bullock
481
505 545
590 650
678
Oil extractors
Above 5 seersa
n.a.
243
66
78 74
40
Less than
5
seersa n.a.
20
212 172 159
76
Shellers
Indigenous
n.a.
n.a.
n.a. n.a. n.a.
175
Power
n.a.
n.a.
n.a.
n.a. n.a. 16
Threshers
Indigenous
n.a.
n.a. n.a.
n.a. 249 n.a.
Power
n.a.
n.a.
n.a.
n.a. n.a. 207
Chaff cutters
Rotary
n.a. n.a.
n.a. n.a.
4,729
n.a.
Power n.a.
n.a.
n.a.
n.a. n.a. 161
Transport
Bullock
8,483
9,862
10,968
12,072
12,695
12,960
Tractor
n.a.
n.a. n.a.
n.a. n.a.
55
Seed
drill/
planter
Bullock n.a.
n.a. n.a.
n.a.
1,135
4,047
Tractor n.a.
n.a.
n.a.
n.a.
n.a.
34
Sprayer/
duster n.a.
n.a. n.a.
n.a. 211 413
n.a. Not available.
a. A seer
equals
about 2.05
pounds.
Source:
Directorate of Economics
and
Statistics,
Agricultural
Situation
in India
(1976),
p.
141;
and
Central
Statistical Organization,
Statistical
Abstract
of India
1975, pp.
57-61.
century,
usually
based on steam
and later
on internal
combustion
en-
gines.
Smaller rice mills have
swept
across
Asia since
the 1950s;
it is
hard
to find
villages
where rice
is still
pounded
by hand.
Thus
mechanical
milling
is even
more widespread
than
mechanical
threshing.
But where
the
green
revolution raised
wages
and increased
harvests
(as in Indian
Punjab,
Philippines,
and
central
Thailand),
the
small threshers
were
rapidly
adopted once
efficient
designs
were
available.
The new
threshers
are now
also
penetrating
into other
South
Asian
regions
(Walker
and
42 Researcb
Observer
1, no. 1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Table
10. Patterns
of
Mechanization
in
China
(thousands)
Four-wheel
Garden
Farm
tractors
tractors
Threshers
Combines
trucks
1957 n.a. n.a.
n.a.
2
4
1962
55 n.a.
n.a. 6
8
1965
73
4
110 7
11
1970
125 78
455
8
16
1975
345
599
1,553
13
40
1979
667
1,671 2,328
23
97
1981
790
2,030
n.a. n.a.
n.a.
Note:
In
1980
there were 560 tractor
trailers
and
36,000
wheelbarrows
with
rubber tires. In
1979
draft
animals,
including
young
stock,
totaled
94,591,
broken down as
follows:
oxen,
52,411;
cows
used for
draft,
558; water
buffaloes,
18,377;
horses,
11,145;
donkeys,
7,473;
mules,
4,023;
and
camels,
604.
Sources:
Ministry
of
Agriculture,
Animal
Husbandry,
and
Fisheries, Agricultural
Yearbook
of
China, 1980 and
China
Academy of
Agricultural
Engineering.
Kshirsagar
1981). As with
earlier American and
European
experience,
mills
and
threshers are
usually rented.
Generalization 9.
The
mechanization
of power-intensive
processing
and
pumping
operations
always
precedes
the
mechanization
of
harvest-
ing
and
crop
husbandry operations and can be
profitable
at low
wages.
Land
preparation.
Unlike
the
power-intensive
operations,
land
prepar-
ation
requires
mobile
sources of
power,
such as
animals,
tractors, or
power
tillers
(hand
tractors). Of
all
land
preparation
operations,
primary
tillage
(breaking
soil,
often
combined
with
turning
its
top
layer)
is
the
most
power-intensive.
It is
also
usually
the
first
use of
a
new
source
of
power.
Investment in
animal-drawn
harrows
occurs
later
and is
usually
much
less
than
investment in
plows. The
widespread
use
of
modern
steel
harrows
in the
United
States
was
delayed
until the
1880s,
roughly
fifty
years after the
massive
shift to
cast
iron
and
steel
plows.
When
tractors
were
introduced,
they
began
to be
used
universally
for
primary
tillage,
while
animals
continued
to be
used
for other
types
of
soil
preparation.
Generalization
10.
Primary
tillage
is
one
of
the
first
operations
to
be
mechanized
when
a
new
source of
mobile
power
becomes
available.
Secondary
tillage
operations
often
continue to
be
performed
by
the
old
power source
for
a
long
period.
Transport.
Carrying
loads is
the
earliest
use of
domesticated
work
animals,
even
preceding
tillage.
Shifts to
animal-drawn
sleds or
carts
follow,
especially
when
marketed
quantities
increase.
The
cart and
plow
Hans
Binswanger
43
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
are the
basic
farmer-owned
implements
of
early
animal-drawn
mechani-
zation
(see
table
11).
When mechanical
power
becomes
available,
it
is soon used for
farm-
to-market
transport.
Early tractors
had
no tires and
in the 1920s
were
rarely
used
for
farm-to-market
transport
in the
United States
or
Great
Britain.
Instead,
mechanizing
farmers
bought
both
tractors
and trucks
at
about
the same time.
That also
happened
in
Mexico
after
1960
(see
table
12).
For on-farm
transport,
American
farmers
continued
to use
horses
well into the
1940s. In
Asia,
where farms
are
rarely
big
enough
to
support
the
purchase
of a
truck,
farm-to-market
transport
is
increasingly
done
by
hired
trucks
or
tractors.
Rubber
tires have
given
tractors
a
strong
comparative
advantage
in most
forms of
transport.
Generalization
11.
Transport,
along
with
primary
tillage,
is one
of
the
first
uses
of
new sources
of
mobile
power.
Where distances
are
long,
trucks rather than
tractors are used
for farm-to-market
transport.
Harvesting.
Without
machinery,
harvesting
is
very
labor-intensive.
Different
crops
vary
widely,
however,
in
the
type
of
labor
required-that
is,
in their
power-
and
control-intensity.
Harvesting
of root
crops
is
probably
the most
power-intensive,
although
it
still
requires
significant
Table
11.
Pattern
of Farm
Mechanization in
Senegal
(thousands)
Year
1950
19S5
1959
1965 1970
1975
Animals
Horses
n.a.
n.a. 98
160
200
210
Asses
n.a.
n.a.
78
147
185
196
Work oxen
n.a.
n.a.
la
1 2
8
Tractors
n.a.
n.a.
0.2
n.a.
0.5
0.4
Plows
0.1
0.6
2
7 8
39
Hoes
0.8
2
4
36
102
219
Harvestors/threshers
n.a.
n.a.
0.1
0.1
0.3
0.1
Carts
Horse
n.a.
n.a.
la
18
23
38
Ox
0.3 3
6
5 5
14
Ass
n.a.
n.a.
0
0.3
6
14
Sowing
machines 11
31
46 94
120
189
Groundnutlifters
n.a.
n.a.
0
6
18
42
n.a.
Not
available.
a.
Figure
corresponds to
1960.
Sources:
Tractors and
harvester/threshers:
Food and
Agriculture
Organization,
Production
Yearbook,
various
issues.
Work
oxen,
1959-65:
World
Bank,
Senegal:
Tradition,
Diversification,
and
Economic
Development
(Washington,
D.C., 1974).
Others: up
to 1955,
Y.
Marie-Saite, La
Culture
attelee
au
Senegal
(Dakar:
Direction de
l'amenagement,
1963); 1959
onward:
Ministere
du
Plan
et de
l'Industrie, Situation
Economique de
Senegal, various
issues.
44
Research Observer
1, no. 1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
control.
At the
other
extreme,
cotton,
fruit,
and
vegetables
require
inten-
sive control:
in
the
case of
apples,
the threat
of
damage
is
so
large
that
their
harvesting
has still
not been
successfully
mechanized.
During the
nineteenth
century many
attempts
were
made
to
develop
harvesting
machinery
in
Europe
and the
United
States
(van
Bath
1960;
USDA
1940).
Reapers
for small
grains
became
widely
adopted
in
North
America after
1850,
with
grass
mowers
for the
dairy regions
following
shortly
afterward. But
in
France
and
Germany
these machines
did
not
make a
substantial
impact
until 1890 or 1900.
This
time
lag
cannot
be
explained by lack of
engineering
knowledge
in
Europe:
the same
coun-
tries
were
using
mechanical
threshers for
virtually
all their
crops and
seed drills
had
already
been
widely
adopted.
The
difference
was
that
labor was more
abundant
in
Europe,
farms
were
smaller,
and
the
har-
vesting
machines were
therefore
not
profitable.
The
United States started
moving
from
reapers
to
wheat binders in
the
1870s
and
to
corn
binders
in
the 1880s. These
changes
coincided
with,
or
even
preceded,
the
development
of
modern
harrowing
technology:
spring
tooth
harrows
and disk harrows.
European
farmers did not
adopt
reaper-binders until
the first decade of the twentieth
century
(Bogart
Table 12.
Pattern
of Farm
Mechanization in
Mexico
(thousands)
Year
1930
1940 1950
1960 1970
Number of
holdings
858
1,234
1,383 1,365
1,020
Work
animals
n.a.
n.a.
3,920
3,476 4,150
Enginesa
n.a.
9
14
18
47
Electric
motors
n.a.
n.a.
n.a.
n.a.
28
Tractors
4
5
23
55
91
Plows
Indigenous
904
925
1,135
1,100
916
Iron
n.a.
720
1,128
1,286
1,301
Harrows
and
cultivators
n.a.
102
240
308
387
Threshers
(fixed)
4b
2b
3b 5
3
Shellers
Engine
n.a.
2
3
5
13
Hand
n.a.
4
5
9
18
Forage choppers
n.a.
2
3
6
6
Carts
106
131
175
211
161
Trucks
4
6
18
40
104
Seed
drills
26
27
60
93
122
Mowers/reapers
8
5
8
10
12
Hay
balers
n.a.
2
3
5
12
Combines
n.a.
n.a.
n.a.
4
7
n.a.
Not
available.
a.
Fixed
and
movable
engines.
b.
May include
some
combines.
Source:
Direcci6n
General de
Estadistica,
Censos
Agricola:
Ganadero y
Ejidal,
decennial.
Hans
Binswanger
45
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
1942).
In
Japan reaper-binders
had a
perceptible impact
only
after
1967,
almost
a
hundred
years
after the
United States
and
a
good
thirty years
after Japan started
mechanizing
pumping, threshing,
and
winnowing in
earnest.
Again,
technological ineptitude
in
Japan
cannot have
been the
cause for
such long delays.
Practical development
of
horse-drawn
harvesting
combines started in
the 1860s
in
California,
where
labor was
extremely
scarce.
By
the
1880s
combines drawn
by
between
twenty-four
and
forty
horses
reaped
ten
to
fifteen hectares
a
day
in California. In the
1890s combines
drawn
by
steam tractors had
a
capacity
of
up
to
twenty
hectares of wheat a
day
(van Bath
1960;
USDA
1960),
but combines did not
spread
beyond
Califor-
nia until
1914.
They
did
not
appear
in Great
Britain
until
1928,
nor in
most
of continental
Europe
until
1935,
and not in
Japan
until
about
1970.
At each level of
mechanization,
machines for
harvesting
maize have
tended to
lag
a few
years
behind those
for small
grains.
Hay harvesting
equipment,
horse
rakes,
and
tedders became
important
during
the
Amer-
ican Civil War
of the 1860s and
remained so
until World War
II.
In
France
in
1892,
hay-raking
machines
had
not reached 10
percent
of
their
1955 peak
number.
Hay
loaders
became
widely
used in the United
States
after 1880 but
did
not
spread
in continental
Europe
until after
World
War
11,
only
to be
quickly replaced
by hay
balers and other
more
sophisticated
machines.
Most of the
animal-drawn
harvesting
machines
derived
their
power
from
horses. Oxen
could not be used
successfully
because
sufficient
power
could be
generated
only
at
the
higher
speed
of
horses.
The
demise
of
the
oxen in
American
and
European
agriculture
was
largely
the
result
of
their
inability
to work
harvesting
machines.
When tractors
became
available, they
could
pull
harvesting
machines
with
only
minor
modifica-
tions.
Nevertheless,
horses did not lose their
comparative
advantage,
even in
harvesting,
for
some
years.
Generalization
12.
Because
mechanization
of
harvesting
is
directly
dependent
on
labor
costs,
it is
rarely
profitable
in
low-wage
countries.
The
higher
the
control
intensity of
the
operation,
the
higher
must
labor
costs
be
to
warrant
using
a
machine.
Crop
husbandry.
Weeding
and
cleaning
of
crops,
fields,
and
orchards
are control-intensive
operations.
In
animal
systems, people
go
on
weed-
ing by
hand
long
after the
introduction of the
plow
and cart-until
rising
wages
make
herbicides profitable.
Wages are so
low in South Asia
that,
except
for
tea
plantations,
it
is still
cheaper to
weed by hand than
to
use
herbicides
(Binswanger
and
Shetty 1977).
Mechanical weeding
between the
rows with
animals
becomes
feasible
only
when line
seeding
is
practiced. Inter-row
cultivation also
tends to be
performed by animals
long
after
tractors are
used for tillage.
46
Research Observer
1,
no. 1 (January 1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Generalization 13.
Crop
husbandry
shifts
to new sources
of
power
only
after
tillage,
transport,
threshing,
and
seeding
have done
so.
Seeding and
planting.
Animal
and tractor-drawn machines
are
capable
of
greater
precision than hand methods
for
only
a few
agricultural
operations,
especially
seeding
and
planting.
Mechanical means of
seed-
ing
may
lead
to
modest direct
improvements
in
yields
and
may
be
attractive in
land-scarce countries with intensive
farming.
The first
seed
drills
were
developed
in
China and
Mesopotamia
in the third
millenium
B.C.
(van
Bath
1960). The
Mesopotamian drill
required
three
workers-
one to
drive the
oxen, one to
put
grain
in
the
hopper,
the
third
to
hold
the drill
steady.
It
was
apparently
possible
to use this
instrument
profit-
ably
only in
the fertile soil of
Mesopotamia,
where
high
yields
could
be
achieved and labor
was abundant. The drill
soon fell into
oblivion.
Between
the
sixteenth and nineteenth
centuries,
farmers in
Europe
tried to
design
better seed drills for small
grains.
Seed drills with
me-
chanical
dribbling
devices were
commonly
used in the United
States in
the
1860s
and
1870s.
In continental
Europe
their
use started
slightly
later,
became
widespread
in the
late nineteenth
century,
and
was fol-
lowed
a decade
later
by
maize drills and
cotton seeders.
The
use
of
seed
drills
similar
to the
Mesopotamian
drill has
been
growing
rapidly in
India
since
1966
(see
table
9).
In
Senegal, where
animal
traction
is
primarily
a
post-1945
development,
seed
drills
have
become
one of
the
most
popular
implements
(see
table
11).
Improved
seed
drills,
with
mechanical
dribbling
of
seeds, are
becoming
popular in
South
Asia and
are
one of the more
successful
machines in
Mexico
(see
table
12).
In all
these
cases
it is
not
labor
saving
which
leads
to
their
success,
but the
improvement
in
yields, the
saving of
seed, and the
ease
of
interculture. For
most
developed
and
developing
countries
for
which
data are
available,
the
spread
of
seed drills is
paralleled
by
the
spread
of
inter-row
cultivators
or,
at an
earlier
stage,
simple
animal-drawn
hoes
or
blade
harrows
for
interculture.
Generalization
14. In
labor-abundant
countries,
seeding
of
grains
tends
to be
mechanized
before
grain
harvesting,
but
the
order
is
usually
reversed
when
labor
is
scarce.
Fertilizer
and
pesticide
placement.
Although
fertilizer
can be
spread
by
hand,
it
produces
higher
yields if
it is
dispensed
precisely.
Thus,
animal-
drawn
machines for
spreading
fertilizer
were
developed
as
fertilizer
use
increased.
Since
fertilizer
was
more
intensively
used in
Europe
in
the
interwar
period,
fertilizer
distributors
were
common
there.
Large
cart-
mounted
barrels
for
spreading
liquid
cow
manure
were
also
widely
used,
as
were
elaborate
pumping
systems. By
contrast,
in
land-abundant
North
America
the use
of
liquid
manure
was
virtually
unknown.
Hans
Binswanger
47
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Liquid
pesticides
cannot
be
applied
without
at least a hand
pump.
Even for
pesticides
in
dust
form,
mechanized
dusters achieve
higher
precision
and reduce waste.
Sprayers
were
developed
at the same time
as
pesticides.
In
France,
for
example,
spraying
carts were
widely
used
in
vineyards
in
1929. But
in
Japan
hand-carried
power
sprayers
for rice
and
other
crops
became
popular
only
with the
development
of a much
wider
range
of
pesticides
after World War
II.
Such
power
sprayers
are now
used
throughout
Asia,
often hired
on
a
contract
basis.
Generalization 15.
The
growth
of
hand
and
power
sprayers
is
driven
by
the
availability
and use
of
pesticides
and is
widespread
even at
very
low
wages. Higher
wages
lead
to the use
of larger
sprayers
which
may
be animal- or tractor-drawn.
Assessment. The selective use of new
power
sources
(particularly trac-
tors)
for
power-intensive
operations
has often been
viewed as a
sign
of
inefficiency.
Since a farmer
makes a
huge
investment
in
a
tractor,
why not
use
it
for all
operations?
U.S. studies carried out
in
the
1920s and
1930s
show
clearly
that there
is
nothing
inefficient in the
selective use of
tractors
for
power-intensive
operations.
As
long
as
agricultural
wages
were
rela-
tively
low, large farmers found it more efficient to
maintain a
tractor
and
truck
along
with
some horses.
Horses
did
virtually
all the
jobs for
which
power
was not the
overriding
input.
Each
power
source
specialized
in
the
tasks
for
which it had
the
greatest
comparative
advantage. Tractors
were
mainly
used for
tillage
and as
power
sources for
stationary machines
such
as
threshers, saws, silo
fillers,
and choppers. The
same
pattern of
tractor
use was
common in
Europe
until
about 1960 and
is
now
common in
South
Asia,
Southeast
Asia,
and
China. The
only
differences are
that
direct
power
takeoff
has
replaced
the
belt and
pulley and that
tractors are
now
more
frequently
used for
transport.
Although
modern
tractors are
more
efficient
than
prewar
ones, wages in Asia
are
much lower
than in
the
prewar
United
States. Asian
countries
are
therefore likely to
make
continued use of
animals
along with
tractors, until
rising wages
make the
animals'
drivers,
and
thus the
animals, too
expensive.
The Speed of
Mechanization
During
the
twentieth
century and
especially
since
World
War
II, the
speed
with
which
farmers
adopt
new
machines has
quickened.
In
Japan,
for
example,
the
number of
motors,
threshers,
and
hullers
increased five-
to
tenfold
between
1939 and
1955.
Power
tillers
grew from
less
than
100,000
to
more
than
3
million
between 1955
and
1975.
Binders,
com-
bines,
and
rice
transplanters
spread
even
more
rapidly
in
the
1970s.
Such
spurts
are
not
unique
to
Japan. Continental
Europe
experienced
many
similar
surges
in
1955-70. In
Taiwan,
after
1968,
it
took
only
about
a
decade to
shift
primary
tillage
completely to
power
tillers.
48
Research
Observer
1, no.
1
(January
1986)
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
Central
Thailand,
in about
fifteen
years
from the late
1960s,
went
over
entirely
to
tractor
tillage
with
locally
designed power
tillers
and
small
four-wheel tractors.
The
adoption
of small
paddy
mills in Southeast
Asia
was
also
very quick.
Rapid change has
not been confined
to the
twentieth
century.
In
the
United States threshers
spread
enormously
from 1830 to 1850
once
satis-
factory designs
were available
(USDA
1940).
The same seems
to
have
happened
in
Europe.
However,
the
speed
of
change
cannot be
captured
by
historical statistics that focus
on
power
sources at a national
level;
it
is
observed
in
operations at a
regional
level. For
example,
the
growth
of
tractors
in
the United States was
spread
over a
fifty-year
period
with
occasional spurts, but once tractors became
available,
they
took over
primary tillage
within
a
much shorter time. Their further
growth
in-
volved
shifting extra
operations
from horses to tractors.
Today
few
farms
in
the Indian
Punjab plow
land
with
animals,
thresh wheat
by
hand, or use Persian
wheels. This is
only fifteen years after
tractors,
threshers,
and
pumps
became
an
important factor
in
Punjab farming.
The
aggregate
Indian data in table 9 hide these
facts because
animals
continue
to be
used
for other work,
even in
Punjab,
and because
many
other
regions
have
not
yet shifted
massively
to tractor
plowing or me-
chanical
threshing.
In
the
case
of
threshers, adoption
cycles have always been
fairly short.
Once
locally adapted
designs
are
available,
the
cost
advantage seems to
be
overwhelming.
For other
machines,
the
explosive growth of the
post-
war
period
must be
understood as
a
response
to
unprecedented
rates of
growth
in
agricultural
wages.
This section
therefore
concludes with two
generalizations.
Generalization 16. Where cost
advantages
are
large
or
change rapidly,
individual
operations
are mechanized
very quickly.
Within smaller re-
gions, adoption periods
are often as short as
ten to fifteen
years.
This
speed of
adoption implies directly:
Generalization 17. In
market
economies, supply
bottlenecks in the
production,
distribution, and servicing
of machines are
rarely a major
cause
of
their
slow
adoption.
The
Process of
Mechanical
Invention
The
previous
sections
imply:
Generalization
18.
Neither
power
sources
nor the
basic
engineering
solutions
for
particular
farming
tasks
are
very
sensitive
to
variations
in
soil
and
weather.
However,
the
power
sources must
be
embodied
in
specific
machines
and
the
basic
engineering
solutions
adapted to
different
Hans
Binswanger
49
This content downloaded from 137.215.6.53 on Wed, 6 Aug 2014 10:07:57 AM
All use subject to JSTOR Terms and Conditions
environments.
Both
agroclimatic
factors
(soil,
terrain,
rainfall)
and
econ-
omic
factors
(land,
labor,
capital,
farm
size,
and materials
available)
require
adaptive
innovation on
a scale that
has
been
vastly
underesti-
mated.
The extent of
adaptive
innovation
required
is
best illustrated
by pa-
tent statistics from
the
United
States,
which
Evenson
(1982)
has
put
together
on
a
regional
basis. For other
developed
countries
and
the
developing
world,
the
investigation
of
innovation
in
agricultural
machi-
nery has been much less
systematic.
Nevertheless,
case
studies,
actual
observation,
and
discussions with
engineers
and
machinery
manufactur-
ers reveal
very
similar
trends.
For
example,
the
emergence
of
a
diversi-
fied
machinery
industry
from
small
shops
is well known
for
the
Indian
Punjab. The Thai
power
tiller
industry
has been well
documented
by
Wattanutchariya
(1981).
Innovations
in the
Philippines
have
been
de-
scribed
by
Mikkelsen and
Langam
(1981).
Generalization
19.
In
the
early phases
of
machinery
invention,
subin-
vention
and
adaptation
are done
almost
exclusively by
small
manufac-
turers or
workshops,
working
closely
with
farmers.
Public
sector re-
search has
contributed little
to
machinery
development,
but
more
to
education. The
contribution
of large
corporations
increases
over
time,
but
continues to
be
largest
in the area
of
engineering operations.
The
reasons for
these
patterns are threefold:
In
sharp
contrast to
biological
innovation,
for
which
public
funding
is
crucial, private
makers of
machinery
can
capture
the
gains
from
their
innovation
by
selling machines. The
innovator's
rights
are
more
protected,
the
more
developed
the
patent
system
is and the
better it
is
enforced.
(For
a full
discussion of
alternative
patent
systems,
see
Evenson
1982.)
*
Because
many
adaptive
discoveries
are
specific
to
particular
regions,
farmers,
blacksmiths,
and
small
firms have an
important
advantage
over
public
research
institutes or
large
corporations.
*
Unlike
biological
or
chemical
inventions,
mechanical
innovation