Speaker
Description
Keywords: welfare; poverty; income; value chains; agrifood systems; resilience; inclusivity; smallholders
Introduction
The analysis of poverty and the policies aimed at its reduction have predominantly been steered by conventional monetary poverty indicators. These indicators differentiate between the poor and the non-poor based on a subsistence threshold, determined by the cost of a basic basket of consumption goods priced at local rates. For international comparison, the now $2.15 ‘dollar-a-day’ concept has guided poverty statistics, establishing a cross-country standard to identify the extreme poor (Gonzalez Sabatino et al., 2023; Ravallion, 2016). While poverty analysis tools have diversified to encompass non-monetary dimensions of poverty and attempt to produce targeting friendly indicators, such as multidimensional measures of deprivation (Alkire et al., 2015), the focus on subsistence has dominated the narrative. This focus has been increasingly challenged by sluggish progress in the eradication of extreme poverty, and more recently by high income inequalities, as evidenced during the COVID-19 pandemic.
The emergence of living wage and income concepts has sought to overcome this shortcoming, defining an adequate remuneration for producers and workers in agricultural value chains as one that would enable attaining decent education, a nutritious low-cost diet, adequate healthcare and healthy housing (Anker & Anker, 2017). In comparison to other monetary poverty measures that do not necessarily reflect the required goods and services inherent in a decent living standard (van de Ven et al., 2021), the concept of living income and its related suite of indicators relies on a normative understanding of a decent life defined according to internationally accepted standards of living conditions but also informed, through a multi-stakeholder process, by elements of local culture and prevailing traditions. This approach aims at guiding policies directed at producers of global value chains, such as cocoa and coffee,
The introduction in recent years in Ghana and the Ivory Coast of a living income differential, a measure to support farmgate prices received by producers of cacao, has raised the question of the best policies and guiding indicators to foster inclusive and resilient agrifood systems in the context of internationally traded agricultural commodities. The Ghanaian and Ivorian living income differential informed a minimum price per tonne of cacao sold to exporters, one that would in theory bridge the gap between the actual income of cacao producers and the income required for those producers to live a decent life – the living income benchmark, LIB (Anker & Anker, 2017). As the LIB embodies the cost for a typical household of obtaining decent education, healthcare, housing and nutrition, the benchmark ultimately exceeds official poverty lines, thus focusing the policy discourse on sustainable and equitable poverty reduction rather than on the subsistence threshold, which does not actively address the risk of poverty traps (Barrett, 2005; Carter & Barrett, 2006)
Objective
The LI methodology, while innovative and promising for addressing income adequacy among agricultural producers, has not been systematically validated following rigorous methods for other welfare indicators (Ravallion, 2016). Our first objective is to assess the statistical reliability of LI indicators and whether they harness utility for targeting purposes. Furthermore, since the ‘living income’ concept was borne from the ‘living wage’ one, conceptualized for wage workers rather than self-employed primary producers, applying the concept to rural, agricultural environments merits assessment of its contextual relevance. We aim to assess whether the LI approach is relevant to the context of small-scale producers. Finally, the cost implication of employing these methods needs specific consideration before scaling up through established data collection processes. This paper aims at addressing these objectives using the Cameroonian cocoa sector as our case study. leveraging a large-scale, representative survey of cocoa producing households in Cameroon.
Methods
We conducted a representative survey of cacao producing households in seven regions of Cameroon to obtain comprehensive data on income, multidimensional poverty, food security, and livelihoods resources and strategies indicators for 4,140 cacao producing households. We estimate the LIB using cost data from nearly 500 sectoral service provider interviews, collected according to two different LI methodologies - the Anker method (Anker & Anker, 2017) and an alternate approach developed by Wageningen University researchers, hereafter referred to as the ‘WUR method’ (van de Ven et al., 2021).
In order to test the relevance, reliability and performance of the LI indicators, we assess their features in terms of the poverty literature axioms of monotonicity, scale invariance, replication invariance, and transferability (Ravallion, 2016). We explore the distributional properties of the LI indicators, following (Kraay et al., 2023) as part of a non-parametric analysis comparing the LI indicators with other established deprivation indicators – the prosperity gap, the multidimensional poverty index and the food insecurity experience scale – to assess errors of inclusion and exclusion. Cost analysis allows us to then quantify the operational implications of each method.
To further evaluate the utility of LI methods for targeting and policy formulation, we identify the factors that contribute to narrowing the gap between incomes and the LIB and test these across the Anker and WUR methods. First, we empirically model the ratio of total annual income to the LIB for cocoa producing households to identify the extent to which specific factors influence the LI gap. Then, given that certain producer types may be closer to the LIB than others, we apply the Oaxaca-Blinder approach (Rodgers, 2006) in order to quantify the extent to which this distance from the LIB is attributed to endowment factors, which are ‘explainable’, versus other unobservable effects, such as prices or discrimination. We define groups according to observable characteristics, namely, gender, age, cooperative membership and production scale.
Findings and conclusions
The findings of this analysis will provide a first comparative study of LI methodologies in relation to other recognized indicators of deprivation. The conclusions of the study will serve not only to guide decision-making regarding the use of these new indicators by policy makers and practitioners, but also provide a new understanding of their potential and limitations when assessed according to a set of conceptual and empirical tools validated by poverty economists.
