transforming west african agriculture through the ...€¦ · change in the shape production...
TRANSCRIPT
Transforming West African agriculture through the development of
mechanization: what public policies?
Keynote Address
Hiroyuki Takeshima ([email protected])Research Fellow
International Food Policy Research Institute, Washington DC
Dakar, Senegal, February 1, 2017
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Outlines
• Patterns of mechanization growths in West Africa and elsewhere
• Private sector in mechanization growth
• Custom hiring service in West Africa - growths and challenges
• Agricultural mechanization policy issues
• Broad agricultural policies and mechanization specific policies, government’s role
• Experiences in Nigeria
• Concluding remark
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Agricultural mechanization has been slow in Sub-
Saharan Africa, including West Africa
Share (%) of land cultivated by
different tools / machineries
Source: Grigg (1984), Mrema et al.
(2008), Takeshima & Salau (2010)5054
8570
4565
89
1525
59
10 20
69
2035
68
2535
71
3221
1122
3025
10
1525
19
1520
17
2037.5
28
2030
23
1725
48
2510
1
7050
22
7560
14
6027.5
4
5535
6
Eastern Africa (97-99)Southern Africa (97-99)
Central Africa (97-99)West Africa (97-99)
(2030)(97-99)
Sub-Sahara Africa (82-84)
(2030)97-99
Latin America / the Caribbean (1982-84)
20301997-99
Near East / North Africa (1982-84)
(2030)(97-99)
Asia (1982-84)
(2030)(97-99)
Developing countries (1982-84)Hand tool
DraftAnimal
Tractor
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
In Asia, tractor use has grown rapidly in the last few
decades
0
10
20
30
40
50
60
70
80
90
100
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020Year
% of areas plowed by tractors in selected countries, year
India
China -
plowing
Indonesia -
rice
CambodiaBangladesh
Nepal
(Lowland)
Nepal
Nigeria
Thailand
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Recent tractor use growth in Asia has been through custom
hiring, rather than ownership growth
0
20
40
60
80
100
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Japan (owner)
In the US or Japan, early tractor uses grew
through ownership growth
*US: ownership of horses/mules declined in
response to tractor ownership => indicates
tractor ownership rather than rental primarily
replaced horses/mules
Recent tractor use growth in developing countries has
been mostly through custom hiring, rather than
ownership growth
Bangladesh (user)
China (user)
Nepal lowland (user)
China (owner)
Bangladesh (owner)Nepal lowland
(owner)
Japan (user)
USA (owner)
USA (horse/mule)
(%)
Indivisible tractor technologies are made scale-
appropriate through custom-hiring in many developing
countries today
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
In Asia, tractor use grew while many workers remained
in agriculture. Same is possible in West Africa
USA (1913)
Australia
France
Japan (1955)
Japan (1970)
China
Bangladesh
Nepal (1995)
Nepal (2010)
Nigeria (2010)
India (1994)
India (2015)
Cambodia
Mali Burkina Faso0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100
Share (%) of area plowed by
tractors
% of employment in agriculture
USA
UK
France
Japan
Australia
China
Bangladesh
Nepal
Nigeria
India
Cambodia
Mali
Burkina Faso
Source: Presenters based on various studies
In the past, mechanization
occurred only after the
agricultural transformation
had occurred
Mechanization today can
both retain smallholders, and
also transform smallholder
farming systems
Would these patterns apply
to West Africa ?
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 7
In Asia, mechanization not only substituted labor, but also
have raised returns-to-scale
→ Intensification (land, fertilizer, seeds etc)
When renting
in tractors
Change in the shape production function
output
Transformative effects on smallholder agriculture
inputs-use intensifications
Specialization
Divisions of labor
Returns to scale: 0.7 Returns to scale: 0.9
→ Intensification (land, fertilizer, seeds etc)
Source: Takeshima (2017).
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
In West Africa, demand for mechanization still largely
depends on farming system evolutionForest fallow
(Slush & burn)
Bush fallow Short / grass
fallow
Annual
cultivation
Multiple
cropping
Fallow period 20 years ~ 6 – 10 years 1 ~ 2 years ~ 1 year No
Cropping 1 – 3 years 1 – 8 years Several A few months
Population ~ 10 / km2
Land clearing Fire Fire
Land
preparation Digging stick
Hoe/digging
stickPlow
Animal-drawn plow
Tractors
Weeding Some Intensive
Use of
animals
Plowing
Transport
Plowing, Transport, Post-
harvest, Irrigation
Labor demand Weeding Land preparation, Weeding, Harvesting
Source: Modified from Boserup (1965), Binswanger and Pingali (1988), Windmeijer and Andriesse (1993)
Population density ↑
Market infrastructure ↑
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
In West Africa, farming intensification has exceeded the
level at which mechanization demand starts growing
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Benin
Nigeria
Gambia
Togo
Burkina
GhanaSierra LeoneNigerGuinea-BissauGuineaSenegal
Liberia
Mali
R-value = harvested area / (arable land + pasture, meadows)
Source: Presenter’s calculations based on FAOSTAT
Farming intensification
often raises demand for
more farm power uses
Example: R-value = 0.33
Switch from long-fallow to
short fallow, and increased
uses of animal tractions
(Boserup 1965;
Ruthenberg 1980; Diao et
al. 2014)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Mechanization patterns are diverse: vary across
regions, and crops (and across countries)
15
19195
0 1
6861
990
0 0
Crops % of crop area using tractors
(Nigeria, 2010/12)
Rice 31
Maize 6
Sorghum 5
Millet 2
Cowpea 4
Ground nuts 3
Cassava 3
Yam 2
Vegetables 1
Source: LSMS-ISA (2010/2012)
Share (%) of land plowed by tractors
or draft animals in Nigeria (2010/12)
Tractors
Draft
animals
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Use intensity of animals is still low in Nigeria: demand
for mechanical farm power (tractors) still insufficient?
Country / regions Reference year Animal tractions (days / per
farms, year)
Source:
Nigeria - North West 2010/12 6 LSMS-ISA
Nigeria – North East 2010/12 9 LSMS-ISA
Bangladesh Early 1990s 90 Mandal & Parker
Japan 1950s 30 Government of Japan
Thailand 1991 15 Pryor (1992)
USA 1930s 100 (including other uses) Jasny (1935)
Note: The figures are derived by the presenters from the original sources, and can differ considerably
depending on the assumptions used.
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 12
CUSTOM HIRING SERVICES OF
TRACTORS IN WEST AFRICA
Private sector has emerged as efficient service providers
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 13
Among various types of custom hiring services,
farmer-to-farmer model is often the most common
N
o
Types of CHS providers Cultivate
own farm
Ownership
of tractors /
power tillers
Sources of
tractors
Share in
Ghana /
Nigeria
1 Specialized service providers No Individual Government
selected
suppliers
Rare
2 Cooperative / Joint ownership Yes Joint Often
government
selected
suppliers
Rare
3 Farmer-to-
farmer
Government-
sourced
Yes Individual Government
selected
suppliers
Relatively
common
4 Farmer-to-
farmer
Market-sourced Yes Individual Market Most common
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 14
Examples in Central Nigeria
Government-sourced
(GS) owners
Obtained tractors only
through government
scheme
Market-sourced (MS)
owners
Obtained tractors
through private market,
private individuals
Tend to be more
efficientEEE
E
EEE
EEEEEE
EE
EEE EEE
E
E
E
EEE
E
EEEE
EEE
E
EE
E
E
E
E
E
E
EE
EEE
EEEEE
E
EE
EE
E
EEE
E
EEEEE
EE
Kaduna
Nasarawa
Abuja
Two types of tractor owners identified based on the
sources of tractors
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 15
Market-sourced owners use tractors more extensively
than government-sourced owners
Source: Presenter’s calculation based on survey.
MS owners operate longer hours, serve more areas than
GS owners
692
499
977*
691
Mean Median
Hours operated, per tractor, year
GS MS
29 41
74
128
GS (mean) MS (mean)
Areas served (ha), per tractor per year
Own farming Hired out farming
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 16
Market-sourced owners realizes higher profitability than
Government-sourced owners
Source: Presenters.
GS MS
Median Mean Median Mean
Total gross revenues per year 5 8 11* 13*
Monetary values of own-farm use 1 2 3 4
Gross earnings from hiring out 4 6 6 9
Payment for operators and fuels 1 3 3 5
Operators 0 1 1 2
Fuels 1 2 2 3
Repairing 1 1 1 1
• MS owners earn significantly more than GS owners
• Revenues are much greater while operators / fuels costs are similar
Benefits from tractor use in the last 12 months (current USD 1,000)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 17
Tractor use highly seasonal but Market-Sourced owners are
more active all-year around
3439
33
6575
111
80
43
24242734
63*5559*
104*
119*
147*
126*
65
44*
57*66*
72*
1 2 3 4 5 6 7 8 9 101112
GS
MS
4340
37
44
61
85
73
52
4236
33
42
63*58*59*
78*
89*92
84
67
4956*
67*68*
38
90
56
20
34*
12*23*
82*
103*
128120*
54
516*
44*51**
Hours of operation per tractor
Mean Median
% operating by month
Month
Substantial seasonality
But MS owners – mitigate seasonality; still find some work in off-season
* = statistically significant difference
between MS and GS
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 18
Market-sourced owners may have knowledge of soil types
and appropriate horsepower
Figure 4. Bulk density of soils in Kaduna and
Nasarawa (darker = heavier soils)Source: ISRIC (International Soil Reference and Information
Centre) (2013).
Kaduna Nasarawa Both
N of
obs
Corr.
Coeffic
ient
N of
obs
Corr.
Coeffic
ient
N of
obs
Corr.
Coeffic
ient
Could select from
a range of
horsepower
138 .196* 26 .577** 164 .241**
All 238 .011 71 .180 309 -.027
Source: Presenters.
Correlation between the tractor horsepower and bulk
density of soil – MS owners (conditional on operating
outside the home LGA)
MS owners who could select from a range of
horsepower
travel more to heavy soil area if they have higher
horsepower tractor
Some indication: MS owners can use tractor
efficiently based on soil type
(no such patterns among GS owners)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 19
Power tiller service providers
Kpong Irrigation Scheme, Ghana
Brief description of Kpong irrigation scheme• One of 22 irrigation schemes in Ghana
• 4000 ha of rice area (2000 ha X two seasons)
• 2500 registered farmers (mostly indigenous farmers)
• Opened in 1959
• Aromatic rice production (Jasmine 85, etc) since 2009
• Yield = close to 6 tons / ha
Power tiller use• 100% of irrigated rice area (4000 ha) prepared by power
tillers
• Private custom hiring with 50 ~ 100 power tillers
• Power tillers introduced by the government, donors, private
companies since 80s
• Growing share of power tillers sourced directly from private
market (20 – 50%)
!
!
!
!!
Kpong
Accra
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 20
Market-sourced power tiller service providers seem
more efficient than government-sourced owners
All
Obtained power
tillers from
private sector
Obtained power
tillers from
government
Categories USD / year USD / year USD / year
Fuel cost 562 626 421
Lubricant cost 184 195 159
Repair cost 424 422 428
Spare parts 570 473 775
Workmanship cost 246 231 280
Operators 513 537 459
Depreciation of power tiller (assuming 5 years) a 713 773 576
Depreciation of implements (assuming 5 years) b 316 359 316
Revenue c 3675 3971 3008
Total area plowed (ha) 23 25 19
Revenue per ha (USD / ha) 158 159 157
Profit 263 542 -311a Based on the (1) power tiller price = $3000; (2) average number of power tillers owned = 1.15. b Assuming that the original value is twice that of current value of implements, and they typically last 5 years.c Imputed revenues for own farm use are included.
Profitability of power tiller service provisions by two-types of power tiller owners (USD / year)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Most tractors are purchased using personal savings;
credit is sometimes provided by tractor importers / retailers
• In Nigeria / Ghana:
• personal savings are generally dominant source of tractor finance
• the dominance of expensive, high horsepower tractors aggravate this
condition
• In Asia, credit is mostly provided by tractor importers / retailers
who also have sufficient know-how of loan management (with
specific divisions dedicated for it)
• Bangladesh
• Transactions costs are high for monitoring payments and insuring default
risks; but the private sector still has incentives as the demand is high
enough to bring profits
• Challenges for some finance mechanisms
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Challenges in West Africa: tractors are large but very few
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70
1952
1958
1960
Japan
2010
Nigeria
1995
2010
Nepal
1994
India
19701974
USA
1945
1930 (US)1984
Indonesia
2010
Indonesia
2008
Bangladesh
1978
China
2010
China
Many countries started out with low
horsepower (10HP) at early stage of
tractor use growth
High average horsepower in low
tractor adoption environment – only
recent phenomena observed in Sub-
Saharan Africa
Indivisibility with tractors is more
serious
Scale of mechanization is definitely
an issue
Average tractor Horsepower
% of area tractor-tilled
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 23
Challenges – mobility of tractor is limited, except small
clusters of long-distance travelers
6 59 12
2326231911 8 9 7
1 2 3 4 5 6 7 8 9 101112
Month
4 4 59
212825
20
95 6 7
20*15*
45*39*
45
23202423*32*
39*
10
Travel alone (N = 79)
Travel in group (N = 11)
6 51012
21232216
11 8 7 7
151110
25
73*
95
62*
84*
1010
53*
15
Travel on tractor (N = 83)
Travel on truck (N = 6)
Group traveler –
- travel further away in
off-season
Truck traveller (10% of sample)
- travel further away in
peak-season
- Travel is generally
confined within 25 km
radius (Euclidean distance)
Average distance (km) away from home district in each
month (Euclidean distance from home district)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
AGRICULTURAL MECHANIZATION
POLICY ISSUES
Labor market
Tractors
Machinery Service providers
Operational
knowledge
Mechanics,
repair service
Spare parts
Land
market
Population growth
Market growth
(food crop,
preferences)
Demand for intensive
land preparation
Demand for transportation
(food, non-food items)
Mechanization demand
Draft animal Mechanization supply Labor
Exchange rate policy
Agricultural R&D and inputs policies
Formal finance, machinery subsidies policy
Informal finance
Infrastructure
policy
Geography, agro-ecological conditions
Crop choice
Farm size
distribution
Education
policy
Crop-livestock
integration
Growth in non-farm economy
Self-learning
Training policy
Information
Public / private
efforts
Trade policies on machineries
Land
policy
Food import
Complex impact pathways of broad economic policies and mechanization specific policies
FuelFuel subsidy
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Agricultural mechanization policy issues
• Mechanization
• is affected more by a broader set of agricultural and economic policies
• is also affected by mechanization-specific policies, but in more limited scope
• Private sector can often meet the demand (if the demand is sufficient)
• Investment in machines
• Provisions of services, spare parts, repairs
• Innovation; engineers in the private sector are often as capable as those in public research
institutions; different from innovations on some other technologies (eg, improved varieties)
• Government can focus more on gathering information and knowledge through
research:
• Effects of broad policies on mechanization, and their mechanisms
• Extent and nature of private sector activities, and where exactly government needs to fill the gap
• Experiences in foreign countries
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Tractor related policies have had mixed effects on
overall mechanization growth in NigeriaPeriods -1972 1973-85 1986-99 2000-12 2013 -
Adoptions (%) – Tractors 1 ~ 5 5 ~ 10 10 8 10 +
Adoptions (%) – Animal traction 3 ~ 4 25 25 +
Animal traction (North) 6 ~ 10 60
Subsidies:- Subsidized distribution of
tractors (fixed amount)- (in the form of interest rate
subsidy)
Large import and subsidized distribution (as many as 3000 tractors / year)
Occasional distribution of subsidized tractors by both Federal and State governments(generally inefficient targeting)
Federal: switch to support for service provider
States: continue subsidized distribution of tractor
Trade policies Generallyliberalized
Import duties for spare parts raised
Devaluation following SAP (8- fold increase in tractor price)Tractors often VAT-exemptImport duties; 5% for spare parts; varied between 0 ~ 25% for tractors
Fuel prices (diesel) Unregulated
Price controlEnhanced domestic refinery capacity
Government control of fuel imports
Diesel - unregulated since 2007Kerosene / gasoline remain subsidized
Industrialization policies (domestic manufacturing)
Joint venture unsuccessful due to poor raw materials
Agricultural R&D Release of farm-power-responsive
varieties (responsive to intensive
tillage – maize, for example) =>
growth of animal tractions thereafter
Reduced support for NARS and reduced farm-power-intensive crops
(limited response to intensive tillage, and demand for farm-power)
Growing food import to fill production shortage
Agricultural R&D Limited effort in public knowledge accumulation (tractor census, agricultural mechanization statistics, etc)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Concluding remark
• Mechanization growth in developing countries in the last 25 years
• widespread mechanization growth is possible for smallholder farmers
• mechanization can grow even when many workers are employed in agricultural sector
• mechanization has transformative effects, rather than simply substituting labor
• In West Africa, overall demand for mechanization is still largely affected by
farming system evolution
• the demand has grown sufficiently high in some areas, but has remained low in other
areas
• Where the growth has occurred, private sector has often emerged as efficient
suppliers of custom-hiring services
• However, accessibility is still constrained
• Tractors are large but few - we need to know more about the optimal size of machines in
West Africa (animals, small pumps, motorcycles are commonly adopted – why not small
tractors?)
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
Concluding remark
• Agricultural mechanization policies
• Agricultural mechanization is affected by broader set of policies, rather than
mechanization-specific policies
• Important to increase overall support for agricultural sector
• CAADP target – allocate 10% of government spending to agriculture
• Agricultural R&D; develop varieties that respond to intensive tillage
• Agricultural R&D on mechanization: gathering information on
• which factors raise demand for intensive farm-power uses
• what the private sector is doing (investment in machines, mechanization service provisions etc)
• situations in other countries (the designs of machineries used, etc)
• Infrastructure development
• rural electrification – lowered the costs of mechanical water pumping in South Asia
• rural road infrastructure
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
References
Diao X, J Silver & H Takeshima. (2016). Agricultural Mechanization and Agricultural
Transformation. IFPRI Discussion Paper 01527.
Hayami Y & VW Ruttan. (1985). Agricultural development: An international perspective.
Baltimore and London: The John Hopkins University Press.
Hayami Y & T Kawagoe. (1989). Farm mechanization, scale economies and polarization: The
Japanese experience. J. Devel. Econ. 31(2), 221-239.
Takeshima H, E Edeh, A Lawal & M Isiaka. (2015). Characteristics of private-sector tractor
service provisions: Insights from Nigeria. Developing Economies 53(3), 188-217.
Takeshima H, R Adhikari & A Kumar. (2016). Is access to tractor service a binding constraint for
Nepali Terai farmers? IFPRI Discussion Paper 01508.
Takeshima H. (2016). Custom-hired tractor services and returns to scale in Nepalese
smallholder agriculture: A production function approach. Agricultural Economics,
forthcoming.
Zhang X, J Yang & T Reardon. (2015). Mechanization Outsourcing Clusters and Division of
Labor in Chinese Agriculture. IFPRI Discussion Paper 01415.
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
THANK YOU !