International Food Policy Research Institute (IFPRI) and Ethiopian Development Research Institute (EDRI). Conference on "Towards what works in Rural Development in Ethiopia: Evidence on the Impact of Investments and Policies". December 13, 2013. Hilton Hotel, Addis Ababa.
Why Teams call analytics are critical to your entire business
The Impact of the Promotion of Row Planting on Farmers’ Teff Yield in Ethiopia
1. ETHIOPIAN DEVELOPMENT
RESEARCH INSTITUTE
The Impact of the Promotion of Row
Planting on Farmers’ Teff Yield in Ethiopia
Joachim Vandercasteelen, Mekdim Dereje, Bart Minten,
and Alemayehu Seyoum Taffesse
IFPRI-ESSPII EDRI
December 13, 2013
Hilton Hotel, Addis Ababa
1
2. Structure of Presentation
Introduction
Set up of the experiment
Estimation strategy
Results
Conclusion
Policy recommendation
2
3. Introduction
Teff is Ethiopia’s most important staple crop in terms of both
production and consumption.
Despite its importance, its yield is remarkably low.
There are several reasons for this; one of the major reasons is
the way farmers are sowing teff.
Reduced seed technologies particularly row planting and
transplanting suggested to improve the productivity of teff.
4. Introduction, cont’d
• New technologies are presumed to be yield
increasing because: 1/ Reduce competition;
2/ Ease weeding; 3/ Lower damage during harvest
5. Introduction, cont’d
Objective of the study: Measure the impact of the
promotion of improved sowing technologies (especially
row planting) on land productivity.
5
6. Set up of the experiment
Design in line with ATA/MoA program, but modifications:
10 AGP woredas were randomly selected in Oromia
4 FTCs were selected in each woreda (40 FTC in total)
In each FTC, 25 farmers were selected (1,000 farmers): 10
controls; 10 row planters, 5 transplanting
At FTC level, experiments on 10 plots conducted by DA
Set-up:
Farmers are randomly assigned to different treatment
groups (control/transplanting/row planting)
All famers (control and treated) received same modern
inputs (fertilizer and improved seed for free)
6
7. Set up of the experiment, cont’d
In total, 3 surveys were conducted:
1. Baseline survey (October 2012)
2. Crop cut (November-December 2012/January 2013)
3. Impact survey (February 2013)
8. Set up of the experiment, cont’d
Problems with implementation:
1 woreda dropped out
25 farmers were not everywhere interviewed
(because of lack of transplanters, etc.)
selection was not everywhere done randomly
Result: two samples of interest
Full sample: both random and non-random
selected farmers
984 farmers from 36 villages
Random sample
537 farmers from 19 villages
8
9. Estimation strategy
Outcome of interest
Yield from farmers plots
Yield declaration by farmers
Yield measurements crop-cut
Yield expectations farmers
Yield from FTC plots
Appropriate econometric methods employed for
different samples and outcome variables
9
10. Results
Effect of row planting on yield measured by crop-cut actual
yield, crop-cut expected yield, and farmers’ declaration
Row planting has a positive - but moderate - effect on teff yield
at the farm level
Table: Impact of row planting (compared to broadcasting) on
yield (%) (*: statistically significant)
Crop-cut
Yield
Rowplanting
Average (control)
(ton/ha)
Crop-cut
expected Yield
Declared
Yield
2
17**
12*
1.1***
1.3***
1.2***
10
11. FTC data
In theory each of the 36 FTC rolled out 10 trials, but because of
implementation problems we have 331 trials.
The effect of row planting: increase of the average yield by 20
percent.
Higher estimates because treatment plots used more
fertilizer, use less seed, and are weeded more
This is on top of better farm management by the Development
Agents (DAs)
11
12. Results, cont’d
Synthesis results:
Row planting has positive effect on yield
However, effects are not always significant
If significant, yield improvements – controlling
for all other inputs - are between 12% and 17%.
On FTC plots, the impact goes as high as 22%
We tried different specifications, but no significant
difference in results
12
13. Conclusions
Both farmer samples as well as FTC data show that row
planting increases teff yields by between 10 and 20%
This increase is significant for yield declarations by farmers;
The effect is much smaller and not significant in crop-cut data
Even if yields do not improve that much, farmers that practice
row planting will still benefit because of lower seed rate.
13
14. Conclusions
This moderate increase in teff yield is in contrast with findings
from demonstration sites and research station trials as well as
farmers’ declared expectations.
Possibly explained by implementation problems, non-optimal
land management, and exaggerated optimism on the potential
of row planting.
As the data is from the first year of roll-out, results can
improve with learning over time.
14
15. Policy recommendation
Most farmers are willing to continue using row planting but
only on small fraction of their teff lands.
Low yield difference and continued practice of traditional
broadcasting implies that:
– more studies have to be done to assess on-farm constraints
– scaling-up new agricultural technologies on a large scale has
to proceed cautiously (allow for learning by farmers)
– Study on labor productivity is crucial
One of the reasons for the gap in yield difference between FTC
plots and the farmers is, inter alia, implementation problems.
More needs to be done to improve outreach to farmers