Adoption of CA practices: evidence of interdependence in plot level farmer technology choice from rural Tanzania. Bekele Shiferaw

Adoption of Conservation Agriculture Practices (CAPs):

Evidence of Interdependence in Plot Level Farmer Technology Choice from Rural Tanzania

Menale Kassie, Bekele Shiferaw, Moti Jaleta, Frank

Mmbando et al

Outline

• Introduction• Objectives• Novelty• Methodology• Data• What we find

(results)• Policy implications

Introduction-1• Declining soil fertility and food insecurity and

poverty are major challenges facing African policymakers today

• The adoption and use of conservation agriculture practices (CAPs) can help overcome these development challenges

• CAPs may offer multiple benefits. But despite substantial initiatives to encourage farmers to invest in CAPs, adoption rates are still low in many countries in SS Africa (Jansen et al. 2006; Wollni et al. 2010; Shiferaw et al. 2011)

Introduction-2

• Relatively little empirical work has been done to examine the socioeconomic factors that influence the joint adoption and diffusion of CAPs, especially conservation tillage, organic fertilizers, legume intercropping and legume rotations (Arellanes and Lee 2003).

• Understanding the determinants of farmers’ choices of CAPs can provide insights into developing strategies for targeting innovations to accelerate diffusion.

Objectives• Determine the extent

of adoption of CAPs among smallholder farmers in SIMLESA project areas in Tanzania

• Assess the interdependence between adoption of different CAPs at the farm/plot level

• Identify land characteristics, household attributes and market and institutional factors that determine farmer investments in CAPs

Why we do this• There is much less research on adoption of

multiple CAPs by the same household; little understanding of complementarities and substitution when farmer invest in alternative options.

• The effect of social networks, market linkages and institutional variables is less understood: – Market access and value chain linkages– Social capital: kinship and local networks– Government effectiveness in services

provision – Biotic and abiotic shocks

Contribution to existing research -2

• To the best of our knowledge, no other study has comprehensively and rigorously analyzed the joint adoption of SAPs in the ESA region. The existing studies in Tanzania (e.g., Mbaga-Semgalawe and Folmer 2000; Isham 2002; Tenge et al. 2004) assessed the determinants of partial technology adoption (fertilizer or SWC), which ignored complementarities and/or substitution effects.

• There are limited adoption studies on conservation tillage, manure use, legume intercropping and rotations in Africa in general and in Tanzania in particular.

Methodology-1

• Jointly analyze the factors that facilitate or impede the probability of adopting CAPs for smallholder farmers in Tanzania

• Multivariate probit (MVP) model– There exist household and field level inter-

relationships between adoption decisions involving various CAP’s

– The choice of technologies adopted more recently by farmers may be partly depend on earlier technology choices --- path dependence

– Farm households face technology decision alternatives that may be adopted simultaneously and/or sequentially as complements, substitutes, or supplements

Methodology-2• Unlike the univarite probit model,

MVP captures this inter-relationship and path dependence of adoption

• Assumes that the unobserved heterogeneity that affects the adoption of one of the CAPs may also affect the choice of other CAPs

• Error terms from binary adoption decisions can be correlated

Data• SIMLESA data

(2010):– 700 farm

households – 1,589 managed

plots– 88 villages– 4 districts

• Data type: detail household, plot and village information collected

• Farming system: maize-legumes

Crop composition: % total cultivated plots allocated to maize and legumes

Crops Karatu Mbulu Mvomero Kilosa Total

Maize 46.9 52.1 61.0 61.4 54.9

Haricot bean 26.6 47.3 14.0 14.4 26.6

Pigeonpea 26.2 0.0 16.6 12.4 13.6

Other legumes 0.4 0.6 8.4 11.7 5.0

Total 100.0 100.0 100.0 100.0 100.0

Total Plots 542.0 535.0 344.0 555.0

Results: descriptive statistics-1

• Definition of Variables and Descriptive Statistics.docx

Adoption of CAPs in TanzaniaMea

nStd. Dev.

Legume intercropping (LI)

Plots received legume intercropping (1 = yes)

0.46 0.50

Conservation tillage (CT)

Plots received conservation tillage (1 = yes)

0.11 0.31

Soil & water conservation (SWC)

Plots received SWC practice (1 = yes)

0.18 0.39

Animal manure(AM)Plots received animal manure (1 = yes)

0.23 0.42

Improved seeds(IS)Plots received improved seeds (1 = yes)

0.67 0.47

Cereal legume rotations(CLR)

Plots received legume crop rotations (1 = yes)

0.17 0.37

Chemical fertilizer (CF)

Plots received chemical fertilizer (1 = yes)

0.04 0.20

Results: descriptive statistics-2

Some explanatory variablesExplanatory variables Mean Std. Dev.

Tenure Plot ownership (1 = owned plot; 0 = rented in plot) 0.89 0.31

Relatives Number of relatives that a farmer have within a village 8.56 15.96

Connectio

ns

Household has relative in leadership position

(1 = yes)0.26 0.44

Market

links

Number of traders that farmer knows

(number)5.69 7.11

ExtensionFarmers trust the skills of extension agents

(1 = yes)0.61 0.49

Pestsdisea

se Pests and disease risk for crops (1 = yes)0.64 0.48

SalariedHousehold member has salaried employment

(1 = yes)0.14 0.35

Gender Gender of household head (1 = male) 0.88 0.33

InsuranceHousehold can rely on govt during crop

failure (1 = yes)0.35 0.50

Rainfalind

ex Rainfall satisfaction index 0.37 0.33

GroupParticipation in farmer coops or association

(1 = yes)0.29 0.46

Empirical Results: Correlation Coefficients for MVP Regression Equations (p-value in parentheses)

SAPs Legume intercropping

Conservation tillage

Manure Legume rotation

Fertilizer SWC

Conservation tillage 0.21(0.00)Manure

0.35(0.00) 0.10(0.26) Legume rotation

-0.3(0.00) -0.16(0.17) -0.39(0.00) Fertilizer

-0.03(0.75) -0.24(0.10) -0.07(0.57) -0.15(0.31)

SWC 0.03(0.59) 0.36(0.00) 0.11(0.09) 0.01(0.91) -0.07(0.52)

Seed 0.50(0.00) -0.02(0.81) 0.13(0.00) -0.17(0.00) 0.42(0.00)-0.03(0.59)

LICTManure LCRCF SWC

Results – CAP adoption factors Expl variables Conservation tillage SWC Legume intercropping Legume rotationsRainfall (---) (--) (+++) (+++)Pest and disease risk (+++) (++) Trusted extension (++) (+++) Connections (+) Group /social network (+++) (+++) Relatives/kins (---) Market linkage (---) Perceived public insurance (+++) (---) (--)Distance to market (---) (---) Distance to extension (---) Timely avail of fertilizer (+++) (++) Salaried hh (yes=1) (---) (-) Gender (male=1) Age (++) (++) Education (+) Livestock assets (+)Farm size (---) (---) (++)Other assets (+++) (-)HH income (+) Owned plot (++) (+++) Plot size (+++) (++) (+++) (---)Distance to plot (+++) (++)Moderately fertile (++) Low fertility (+++) (--)lModerate slope (++) (+) (--)Steep slope (---) (++) (++) Moderate depth (+)

Results: CAP adoption factors (2)Variable

Animal Manure

FertilizerImproved

seedRainfall Pest and disease risk (+++) (--)Trusted extension (+++)Group /Network (+++) (+++) Market linkage (+++) (+)Distance to market (+++) Time avail of fert Salaried hh (yes=1) (--) Farm size (+++) Age (---)Education (+)Livestock assets (+++) (++) (+++)Farm size (--) (---)Other assets (++) HH income (+++)Owned plot (+++) (+++) Plot size (+++) (+++)

Distance to plot (--) (+++) Low fertility (++) Gentle slope (+) Steep slope (+++) (-)Moderate depth (++)

District level effects

Districts Conservation tillage

Legume intercropping

Legume rotations

SWC Animal manure

Fertilizer Improved seed

Mbulu   (+++) (---)   (+++) (-) (---)

Mvomero   (---) (---) (---) (---) (+++) (---)

Kilosa (---) (---) (---) (---) (---)   (---)

Reference is Karatu district

Effect of CAPs on Crop Production Kolmogorov-Smirnov Statistics Test

SAP type Distribution

Legume intercrop (LI)0.2444

(p = 0.000)***

Animal manure0.2474

(p = 0.000)***

Improved seeds0.2762

(p = 0.000)***

Chemical fertilizer (CF)0.1471

(p = 0.317)

Soil and water conservation (SWC)

0.0615

(p = 0.440)Conservation tillage (CT) 0.1059

(p = 0.087)*Legume crop rotation (LCR)

0.0522

(p = 0.636)

0

.2

.4

.6

.8

1

Cum

ulat

ive

Prob

abilit

y

0 1000 2000 3000Net value of crop production

Without legume intercrop With legume intercrop

Figure 1. Impact of legume intercrop on net value of crop production(' 000 TSh/acre)

Empirical results: MVP reults-1

• Production risk: The probability of adoption of CT, SWC and LI is more common in areas and/or years where rainfall is unreliable (in terms of timelines, amount, and distribution)

• Extension - The quality of extension positively influence adoption of CT, SWC, and improved seeds.

Empirical results: MVP reults-2

• Markets -The probability of adoption of capital-intensive practices: improved seeds and fertilizer, increase with enhanced value chain linkages (through links with traders).

• Rural institutions -Participation in rural institutions (groups, networks) enhances adoption of CAPs (LI, SWC, animal manure and fertilizer).

• Public insurance - expectation of public safety nets seems to reduce legume intercropping but increase SWC.

• Off-farm income seems to be negatively associated with CAP investment (poverty or specialization effect?)

Empirical results: MVP reults-3

• Land tenure influences adoption of SWC, CT, & animal manure, which is more common on owner-cultivated plots than on rented in (or borrowed) plots.

• Labor - availability of family labor is positively associated with adoption of manure in crop production

• Livestock also has positive effect on adoption of improved seeds, fertilizer and legume rotations.

Empirical results: MVP reults-4

• Farm equipment ownership has a positive and significant effect on adoption of CT and fertilizer.

• Farm size - Households that own less land are more likely to adopt CT, LI, fertilizer and improved seeds; but households with more land practice legume rotations.

• Plot characteristics are important determinants of CAP choice. Example - farmers are unlikely to adopt CT, SWC, LI and improved seed on small plots. SWC common on poor soils with gentle/steep slopes .

Conclusion• Plot level interactions are important in

identifying suitable CAP combinations for specific environments.

• Policies that properly target CAPs based on agro-ecology and are aimed at organizing small-scale farmers into associations, improving market linkages, education, and enhancing skills of civil servants can increase adoption.

• Economic benefits from CAPs vary – good practice to identify options that offer relatively quick benefits to farmers.

• Future analysis needs to examine the productivity, risk, environmental and welfare implications to particular CAPs and combinations of sustainable agricultural practices.