2017 l VOLUME 8 l ISSUE 4

VICE PRESIDENCY FOR ECONOMIC GOVERNANCE

The unintended consequences of agricultural input

1| Introduction
Modern agricultural inputs such as inorganic fertilizer and human health (e.g., aflatoxin), increase labor requirements and
pesticides potentially help farmers boost productivity severely limit yield amount and quality. Herbicide use reduces
significantly, a goal critical to structural transformation and the drudgery associated with hand-weeding, which may
poverty reduction, particularly in regions like sub-Saharan Africa improve quality of life, and decrease energy expenditure,
(SSA). There is a strong, causal relationship between the use of physical hardship and risk of injury. Improved yields likely
modern agricultural inputs and crop yields and, subsequently, translate into improvements in human health of farming
economic growth (McArthur and McCord 2014). This new households. This can take place via nutrition and/or income
empirical evidence builds upon a well-theorized literature and effects in that farming output can improve nutrition through
brings into focus the drivers of agricultural productivity growth as direct consumption, or through income that can indirectly
a prerequisite for structural change in SSA and elsewhere. But improve human health if households use it to purchase
the use of modern agricultural inputs may put human health and nutritious food, health care and/or preventative care. Addition-
the surrounding environment at risk, thereby decreasing net ally, consumers benefit from increased yields through increased
growth in productivity and well-being in the short and longer food supply, which should reduce prices in areas not well
run. These unintended consequences may be most true of integrated into national and global food markets. Additionally,
pesticides – like insecticides, fungicides, and herbicides – many release of labor from manual agricultural tasks may contribute
of which are known to be toxic to humans, particularly when to more vibrant and economically diverse rural areas. Further
over-applied or used without appropriate protective equipment. afield, controlling pests on export crops can mean the geo-
This brief describes the tradeoffs between productivity benefits graphical containment of pests that could potentially negatively
and human health costs of pesticide input use as observed in affect other countries’ environments and farming systems.
SSA. Prior work on this question is limited to small sample
studies that focus on cash crop systems. Here we present the Pesticides are often toxic to humans. Depending on type and
first broad, cross-country, and nationally representative analysis class, pesticides can contribute to neurological, respiratory,
of productivity and health impacts of pesticide use in developing immunologic, and reproductive health problems (Weisenburger
country agriculture across multiple crop types. 1993) as well as cancer, immune system damage, and
increased risk of sickness in the short term (Culliney et al.
The prospective gains from pesticide use are considerable. 1992). Harmful encounters with pesticides can occur in a
Pesticides reduce incidence of harmful pests—insects, fungal number of situations. Most directly, farmers or other agricultural
pathogens, weeds, etc.—which can both directly impact laborers applying these chemicals to crops risk contact via

1 Megan Sheahan ([email protected]) is with Precision Agriculture for Development. Christopher B. Barrett is with the Cornell SC Johnson College of
Business. Casey Goldvale s with the Georgetown Center on Poverty and Inequality.

Disclaimer: The findings of this Brief reflect the opinions of the authors and not those of the African Development Bank,its Board of Directors or the
countries they represent.

1
 AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

exposed skin and eyes, both of which can absorb chemicals at pesticides of one type or another. In Nigeria and Ethiopia,
potentially toxic levels, or through ingestion via the mouth and pesticide use is much higher at 31% and 34%, respectively.
nose. Beyond the time of application, contact with chemical Herbicides are the most commonly used pesticide, followed
residues during other agricultural tasks (like weeding, thinning, by insecticides. Figures 1 and 2 display pesticide use in
and harvesting) can also be problematic. Limiting exposure is Ethiopia and Nigeria, respectively, disaggregated by pesticide
possible by wearing protective equipment and other strategies type. The fact that pesticide use is higher than commonly
to keep the chemicals away from the body. Protective acknowledged underscores the importance of exploring the
equipment use is low in SSA, suggesting considerable scope tradeoffs associated with pesticide use in these countries.
for contamination (Ajayi and Akinnifesi 2007; Banjo, Aina, Rije
2010; Mekonnen and Agonafir 2002). Pesticides are associated with agricultural productivity.
Pesticide use is strongly and consistently associated with
Pesticide exposure is not limited to agricultural laborers. Other greater agricultural output value in Ethiopia, Nigeria, Tanzania,
members of farming households—including children—are and Uganda. Plots where pesticides were used see total
likely to walk through or play in fields with chemical treatment, harvest values increasing by $32-$85 on average, depending
especially those located near dwellings. The storage of pesti- on the country. Alternative estimates suggest that there is a
cides, especially in open containers, in close proximity to 33% increase in harvest value in pesticide-using plots across
where household members congregate, eat, or sleep is all countries.
another way for household members to come into contact
with harmful substances. Rural agricultural households often Pesticide use is associated with higher health care costs. We
reuse sturdy, plastic pesticide containers for water, grain, oil, find that pesticide use is correlated with higher levels of
or other consumables, putting family members at risk for expenditure on health care as a result of sickness in Tanzania
ingesting pesticide residues. Williamson (2003) notes that and Uganda. We rule out the possibility that this relationship is
over three-quarters of all pesticide poisoning cases reported driven by wealthier farming households’ ability to purchase
to partners of the Pesticide Action Network in Benin and more pesticide inputs and more health care, simply by virtue
Senegal were related to food and drink contamination, not to of their higher income levels. On the contrary, it may be the
exposure on fields. Though beyond the scope of this study, case that health expenditures underestimate the incidence of
potential for such contamination extends beyond agricultural illness related to pesticide use, particularly in settings where
households to the consumers of agricultural products. access to health care is limited or households are liquidity
constrained, and therefore will only seek costly health care in
After application, pesticides can also pollute the environment extreme cases. If this is true, then the relationship between
from which rural households critically depend and derive pesticide use and health expenditure can be considered a
livelihoods, indirectly affecting human health. Pesticides used in lower bound of a possibly larger relationship.
high amounts or applied at inappropriate times, such as just
before rainfall, contribute to chemical run-off and the Pesticide use is linked to other variables implying poor health.
contamination of drinking water for the surrounding rural First, we find that households that use pesticides are 4-9%
population. Pesticides can also damage agricultural soils more likely to have missed work due to illness in the previous
through the degradation of beneficial soil microorganisms and 1-2 months than households that do not use pesticides in their
the sorption or binding of important organic or mineral farms. This result is remarkably consistent across the three
components (van der Werf 1996); poor soils will inevitably lead countries for which these data are available (Ethiopia, Nigeria,
to lower harvests. and Uganda). Second, we find that Ethiopian, Nigerian, and
Ugandan households that use pesticides report are more likely
Data now exists to study these relationships in SSA. Given to have experienced a recent illness among their members and
the potential for countervailing effects of pesticide use, we to have made more visits to a health worker. Tanzanian
use the Living Standards Measurement Study Integrated pesticide-using households also report more visits to a health
Survey on Agriculture data from four countries—Ethiopia worker. For income and access reasons, we would not expect
(2011/12 and 2013/14), Nigeria (2010/11 and 2012/13), that all individuals suffering from an illness – related to pesticide
Tanzania (2008/09, 2010/11, and 2012/13), and Uganda exposure or otherwise – would visit a health worker for
(2009/10, 2010/11, and 2011/12)—to explore the relation - treatment or advice, and especially not if they consider the
ships between human health and agricultural productivity with negative health effects normal or routine (Banjo et al. 2010).
respect to pesticide use. The findings below should not be However, the consistency of the pattern is striking and is a
interpreted as causal, nor as comprehensive estimates of cause for concern.
such effects, but rather as a jumping-off point for further
analysis on this very important topic. Herbicides underpin pesticide-health relationships. Of the
three types of pesticides that appear in our data (herbicides,
insecticides, and fungicides), preliminary evidence shows that
2| Key findings herbicides account for all of the relationships above in Ethiopia
and Nigeria. Recall that herbicides are the most commonly
Pesticides are more widely used than commonly acknow- applied pesticide, particularly on staple crops such as teff and
ledged. Sheahan and Barrett (2014) show that 13% and 15% wheat (Ethiopia) and maize and rice (Nigeria), but that prior
of agricultural households in Tanzania and Uganda use work on the health effects of pesticide use focuses on cash

2
 AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

crops such as cotton. When we limit our analysis to a bundle point to the importance of conveying good and accurate
of staple crops, we continue to find the positive relationships information about the risks of pesticide use, particularly
between pesticide use and negative health implications, overuse, and doing so using participatory methods
suggesting that any these relationships cut across crop types. (Dasgupta, Meisner, Huq 2007). At the same time, more
judicious use of pesticides due to greater knowledge about
Gender matters for pesticide-health relationships. Finally, the ideal application conditions or amounts could also have
association between pesticide use and negative health positive crop output implications, implying even further
outcomes appears to be stronger for males than for females, benefits to household productivity or net income levels,
though the mechanism is unclear given that prior work using especially if reducing chemical use cuts down on input costs
the same data finds no discernable gender difference in or enables farmers to preserve the natural environment upon
agricultural activities that might be associated with different which their livelihood is based. The development community
levels of pesticide exposure (Palacios-Lopez, Christiaensen, should remain cautious about how to couple access to
Kilic, 2015). This points to a need for further investigation modern inputs with the requisite knowledge and precautions
along a gender dimension. to keep rural households healthy and productive, especially as
rising wages will naturally coincide with rising agricultural
productivity as structural transformation renders labor-saving
3| Conclusions and policy implications pesticides more attractive.

While we cannot say for certain that pesticides as they are We offer these relationships as a call to other researchers to
currently used in SSA cause health problems, the consistency better understand the decision making and behavior that
in our estimated correlations – across samples, specifications, underpin our results using more tailored questionnaires to
and estimators combined with intuition and theory – suggests help answer the obvious follow on questions. Are farmers
that much more attention should be directed towards operating without full knowledge of the potential human health
understanding the causal link and, where it truly exists, the costs? Or are the costs theoretically known, but farmers
extent to which it might be mitigated with better policies or unable to make the link for themselves that the sickness
programs to promote farmer awareness of the human health occurring within their households may be driven by the use of
consequences of pesticide application rates. Our results are pesticides on farm and stored within the family dwelling? Or,
consistent with a stylized model in which trade-offs exist and perhaps even worse, are they fully aware that household
information gaps for farmers might naturally lead to dangerous sickness is related to pesticide use and continuing to apply
exposure to toxic pesticides. Our findings lay a foundation for them at the expense of household members’ health?
more detailed exploration of whether farmers accurately
understand the prospective health effects of pesticides use Where pesticide use may undermine human health status,
and incorporate that knowledge into both chemicals handling more focused and intensive investigation of that adverse
and storage practices and application rates. relationship will inform discussions about potential extension
and regulatory programming, both within the agricultural and
In light of our empirical results, one might expect that public health arenas. Forsaking the health, and thereby
extension efforts, either public or private, aimed at informing productivity, of the very individuals who will carry out the
farmers about the potential negative human health effects of structural transformation still yet to truly unfold in rural Africa
pesticide use could promote optimal use of pesticides. while promoting the use of yield-enhancing inputs may prove
Economists who have studied pesticide use in other contexts unwise.

3
 AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

References

Ajayi, O.C., & Akinnifesi, F.K. (2007). Farmers’ understanding of pesticide safety labels and field spraying practices: a case
study of cotton farmers in northern Côte d’ Ivoire. Scientific Research and Essay, 2(6), 204–210.

Banjo, A.D., Aina, S.A., & Rije, O.I. (2010). Farmers’ Knowledge and Perception Towards Herbicides and Pesticides Usage in
Fadama Area of Okun-Owa, Ogun State of Nigeria. African Journal of Basic and Applied Sciences, 2(5-6), 188–194.

Culliney, T., Pimentel, D., & Pimentel, M. (1992). Pesticides and natural toxicants in foods. Agriculture, Ecosystems &
Environment, 41, 126-150.

Dasgupta, S., Meisner, C., & Huq, M. (2007). A pinch or a pint? Evidence of pesticide overuse in Bangladesh. Journal of
Agricultural Economics, 58(1), 91-114.

McArthur, J.W., & McCord, G.C. (2014). “Fertilizing growth: Agricultural inputs and their effects in economic development.”
Global Economy and Development Working Paper No. 77. Brookings Institute: Washington, DC.

Mekonnen, Y., & Agonafir, T. (2002). Pesticide sprayers’ knowledge, attitude and practice of pesticide use on agricultural farms
of Ethiopia. Occupational Medicine, 52(6), 311–5. Retrieved from http://www.ncbi. nlm.nih.gov/pubmed/12361992.

Palacios-Lopez, A., Christiaensen, L., & Kilic, T. (2015). “How Much of the Labor in African Agriculture is Provided by Women?”
Policy Research Working Paper No. 7282. World Bank: Washington, DC.

Sheahan, M., & Barrett, C.B. (2014). “Understanding the Agricultural Input Landscape in Sub-Saharan Africa: Recent Plot,
Household, and Community-Level Evidence.” Policy Research Working Paper No 7014. World Bank: Washington, DC.

van der Werf, H. (1996). Assessing the impact of pesticides on the environment. Agriculture, Ecosystems, & Environment, 60,
81-96.

Weisenburger, D.D. (1993). Human health effects of agrichemical use. Human Pathology, 24(6), 571–576.

Williamson, S. (2003). “Pesticide Provision in Liberalized Africa: Out of Control?” Agricultural Research and Extension Network
Paper No. 126. Overseas Development Institute: London, UK.

4
AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

Figure 1 Map of agro-chemical use in Ethiopia Y1 (2011/12) cross section

Source: Authors’ calculations using the Living Standards Measurement Study Integrated Survey
on Agriculture cross section for Ethiopia 2011/12.

5
AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

Figure 2 Map of agro-chemical use in Nigeria Y1 (2010/11) cross section

Source: Authors’ calculations using the Living Standards Measurement Study Integrated Survey
on Agriculture cross section for Nigeria 2010/11.

6
AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

7
AEB 2017 l VOLUME 8 l ISSUE 4 l VICE PRESIDENCY FOR ECONOMIC GOVERNANCE AND KOWLEDGE MANAGEMENT

DESIGN AfDB PCER / 2017