Fertiliser subsidy and agricultural productivity in Senegal

• Published date: 01 September 2017
• DOI: http://doi.org/10.1111/twec.12487
• Author: Abdoulaye Seck
• Date: 01 September 2017

ORIGINAL ARTICLE

Fertiliser subsidy and agricultural productivity in

Senegal

Abdoulaye Seck

Department of Economics, Cheikh Anta Diop University, Dakar, Senegal

1

|

INTRODUCTION

Fertiliser subsidy is one the most difficult agricultural policy decisions faced by governments across

Africa. Its complexity as well as its many challenges has not been reduced by the theoretical and

empirical literature that tends to suggest opposing results when it comes to the ability of the subsidy

programme to raise the productivity of the farming activity. The change in the relative input prices

induced by the subsidy should weaken the budget constraint faced by farmers, expand the technical

space, allow additional investment and lead to more efficient, once-unreachable, input combinati ons.

This efficiency-ladder effect, also referred to as a “crowd-in”effect, is synonymous with greater pro-

ductivity (Mason & Jayne, 2012). Furthermore, even if the production technique does not change,

there could be a productivity gain in an input-oriented sense, for the farmers can now reach the same

level of output with reduced input costs. Additional arguments relate to the possibility to overcome

missing or imperfect markets for farmers, such as the credit market (IFDC, 2003) or the insurance

market (Donovan, 2004), as well as the possibility to correct negative externalities (Gladwin, Randall,

Schmitz, & Schuh, 2002). On the other hand, opposing arguments suggest that fertiliser subsidy can

lead to a “crowd-out”effect by turning farmers away from more sustainable, profitable and promising

land-use practices such as organic matter, minimum tillage and low-input agroforestry (Donovan,

2004). The subsidy could also lead to a possibility of “leakages”by providing incentives for farmers

to resell at higher prices the fertilisers they obtain for next-to-zero price (Mason & Jayne, 2012).

On the empirical side, most of microlevel studies have been focused on the impact of fertiliser

subsidy on aggregate measures such as national food production or broad measures such as total

production or yield (Jayne and Rashid, 2013). For instance, Dorward and Chirwa (2011) have

shown that heavily subsidised fertilisers, in conjunction with the large-scale disbursement of hybrid

and composite seed, have contributed to significant increases in national maize production and pro-

ductivity in Malawi. Similarly, Ricker-Gilbert and Jayne (2010) find fertiliser subsidy is associated

with increased production, with an important dynamic effect. World Bank (2010) also sugges ts a

positive effect of fertiliser subsidy on maize production in Zambia, which corresponds to 89% of

growth in output as a result of the programme, mostly due to higher yield, which has increased by

50%. On the other hand, Fearon, Adraki, and Boateng (2015) provide empirical evidence that the

Ghanaian fertiliser subsidy programme has not yielded a significant increase in total crop output.

In the face of less conclusive literature, this paper aimed at assessing the impact of public

support to agriculture on the productivity of farmers. More specifically, it empirically provides

answers to the following research questions: (i) in the specific context of the Senegalese

DOI: 10.1111/twec.12487

World Econ. 2017;40:1989–2006. wileyonlinelibrary.com/journal/twec ©2017 John Wiley & Sons Ltd

|

1989

agricultural sector, what are the farmers’response mechanisms to the incentives brought about by

the subsidy programme? (ii) Do these mechanisms translate into greater efficiency?

This research adds to the literature by first considering an alternative measure of perfor mance,

namely productive efficiency, which goes beyond simple productivity, and which is obtained

through a non-parametric approach in conjunction with a regression analysis based on an endoge-

nous switching regression model. As opposed to the traditional parametric regression of agricul-

tural production processes, non-parametric approaches such as data envelopment analysis are able

to better capture the heterogeneity of productive units, efficiency-wise, instead of just assuming

that the same technology used by all of them. In addition, the research focused on a very particu-

lar agro-ecological region where irrigation is highly prevalent (thanks to the proximity of the

Senegal River) and where the government has heavily invested in terms of infrastructure (irriga-

tion system, roads, etc.), land preparation and many other extension services. The paper asks

about the effectiveness of the programme in delivering the efficiency outcome in this particular

context.

Any success of the fertiliser subsidy programme in terms of increased farming efficiency woul d

be very beneficial to the country where close to half of the labour force contributes to a meagre

14.8% of gross domestic product (GDP) in 2014. This makes the average farmer more than six

times less productive than her average counterpart in the rest of the economy. The low productivity

of the sector greatly contrasts with its somewhat unrivalled potentials for improving economic and

social conditions. In effect, a more productive agriculture would contribute to shrinking the enor-

mous food gaps that mainly originate from low domestic production relative to the ever-increasing

domestic demand. In addition, by increasing the domestic components of the food markets and

reducing the structural dependence on imports, greater productivity in agriculture would help

reduce the vulnerability to foreign shocks, particularly those related to prices. The 2008 world food

crisis has contributed to a sharp increase in domestic prices in West Africa. In the case of rice,

local prices have increased by between 50% (Mali and Benin) and 100% (Senegal), with the con-

cerns that it might lead to a worsening of living standards, especially among the poorest segments

of the population.

1

An increased productivity in the agricultural sector would help reduce the adverse effect of for-

eign shocks and improve food security by increasing the availability of agricultural products, as

well as their affordability, thanks in part to more favourable price dynamics. Furthermore, it would

provide increased income and economic opportunities to a large majority of the population that

depends directly or indirectly on the sector for their livelihood, particularly in the rural sector,

thereby reducing poverty and inequality.

The study is based on recent farm-level data collected in the agro-ecological regio n of the

Senegal River Valley. The survey collected detailed information on more than 180 farmers. The

data reveal that about half of farmers benefit from the subsidy programme. The empirical

approach is first based on a non-parametric approach known as the data envelopment analysis

(DEA) which estimates technical efficiency of the farming units. Then, a variant of switching

models known as an endogenous binary-variable model is developed to relate efficiency and vari-

ous covariates, chief among them being whether the farmer has benefitted from the fertiliser sub-

sidy programme. This model addresses the potential endogeneity and self-selectivity issues

associated with the use of subsidised fertiliser, and views the latter as an endogenous binary-treat-

ment process.

The remainder of the paper is organised as follows. Section 2 details the methodological

approach. Section 3 describes the data. Section 4 presents the results and their interpretati on.

Finally, Section 5 offers concluding remarks.

1990

|

SECK