Agricultural Productivity, Poverty Reduction and Inclusive Growth in Africa: Linkages and Pathways

Poverty in Africa is primarily rural concentrated, about 75% of the poor population live in rural areas and draws their livelihood and food from agriculture. The Sub-Saharan African region is home to more than quarter of a billion people living in extreme poverty, with the Eastern and Southern Africa having the world’s highest concentrations of poor people. The renewed focus on the poverty reducing potential of agricultural productivity accentuate from the fact that the incidence of poverty in Sub-Saharan Africa is increasing faster than the population. The study examined the effect of agricultural productivity on poverty reduction in Africa using the dynamic panel data approach estimated using the System-GMM technique for the period 1991-2015. The conceptual framework of the study identified three main linkages via which agricultural productivity translates to poverty reduction; this include: i. income empowerment, ii. Market expansion, and iii. Sustenance enhancement. The empirical result suggests that agricultural value added per worker contributes significantly to reducing rural poverty in Africa. On the other hand, food production index and GDP per capita were more important factors in curbing urban and dollar poverty implying that non-farm poor tends to have a large food marginal propensity to consume (MPC). The insignificance of GDP per capita in dwindling rural poverty reflects that the reality that growth in other sector does not influence the livelihood of the rural poor farmers due to its subsistence nature. Finally, domestic credit to private sectors and institutions were significant in reducing all categories of poverty, with largest impact on rural poverty. It implies that development programmes targeted at enhancing agricultural productivity should encompass strategies for accessing credit in order to boost the asset base of rural farmer for a large scale commercial production. Also, appropriate macroeconomic policies and institutional quality needs to be enhanced to boost provision of social services, equitable land and credit access.


Introduction
Most of the world's poor are rural, depending on agriculture for livelihood; thus the linkage between rural poverty and agriculture is necessarily a close one. Poverty is predominantly rural both in its absolute and relative measure in Africa. In Sub-Saharan Africa, more than 65% of the population are rural, out of which 56% depend on agriculture for their for their livelihood. Also, agriculture in the region is largely subsistence and production is concentrated in low-value food cropsaccounting for more than 70% of the regions agricultural output. This makes rural poverty transgenerational in its form due to limited asset base, weak or non-existent market linkages and lack of access to financial services. The renewed commitment to agriculture at national and global level intensified in 2009, following the fact that for the first time in history, the number of hungry people in the world surpassed 1 billion. This was largely as a result of the earlier food and financial crises.
The recent comeback by the development cooperation in exploring the dynamics of agriculture and rural growth promotion reveals some signs of a reversal in the long-term neglect of agriculture. Also, considering the rising statistics of poverty with focus on the sector and space where the poor are employed and lives respectively; policy makers have come to realize that poverty reduction in developing countries is achievable only if development efforts are targeted at agriculture.
The MDG set in with a target of halving the number of people living in extreme povertyproportion of people whose income is less than 1US$ per day and suffering from hunger.
Exceptional progress 1 were made in some developing countries, however, a number of countries fall short and up to 1billion will likely remain destitute by 2015. For over 30 years on, those living in developing countries depending on agriculture for a living are typically much poorer than those working in other sectors of the economy, which usually represent a significant share of the population. In the words of Schultz (1979) 2 , effectiveness in addressing poverty and 1 The share of undernourished people in the region's population fell from 35% (1990/92 MDG base) to 32% (2001/03). Countries like Ghana and probably Gabon have already met MDG goal on undernourishment. Most success stories correlated with agricultural production growth. 2 Theodore Schultz 1979 Nobel price acceptance speech -"most of the people in the world are poor, so if we knew the economics of being poor we would know much of the economics that really matters. Most of the world's poor people earn their living from agriculture, so if we knew the economics of agriculture we would know much of the economics of being poor" alleviating the standard of living of the World poor requires an adequate knowledge of the economics of agriculture.
Theoretical reviews identify the linkages between agricultural productivity and poverty reduction. Available evidence suggests multiple pathways through which increases in agricultural productivity can reduce poverty; this include real income changes, employment generation, rural non-farm multiplier effect and food prices effects. Likewise, DFID (2004) outlines four channels through which agricultural productivity reduces the incidence of poverty, comprising i. direct impact of improved agricultural performance on rural incomes, ii. Impact of cheaper food for both urban and rural poor, iii. Agriculture's contribution to growth and the generation of economic opportunity in the non-farm sector, and iv. Agriculture's fundamental role in stimulating and sustaining economic transition 3 . Also, Bresciani and Valdes (2007) posits that labour market expansion, rising farm income and declining food prices are the three key channels that link agricultural growth to poverty. Thirtle et al., (2001) concluded that agricultural productivity growth has a robust and consistent impact on poverty for all productivity measures.
However, Schneider and Gugerty (2011) identified limited initial asset endowment, barriers to technology adoption 4 and constraints to market access as inhibiting the ability of the poor to participate in the gain from agricultural productivity growth.
The existing empirical literature and theoretical researches addressing the subject matter suggests that agricultural income growth is more effective in reducing poverty than growth in other sectors due to two factor. First, because the incidence of poverty tends to be higher in agricultural and rural populations than elsewhere. Secondly, most of the poor live in rural areas and a large share of them depend on agriculture for a living (Cervantes-Godoy 2010; Christiaensen and Demery, 2007;Ravallion and Chen, 2007). Christianensen and Demery (2007) justified this claim by illustrating empirically that benefits accruable from agricultural growth can be easily obtain if the growth occurred where they are located, implying that the contribution of economic growth to poverty reduction differs across sectors. The underlying reasoning hinges on the assumption that market differentiation, remoteness or political economy consideration makes it difficult to transfer income generated in one geographical location or sector to another.
Similarly, Ravallion and Chen (2007) opined that the poverty reduction impact of agricultural growth tends to be four times greater than growth in secondary and tertiary sectors. On the other hand, Warr and Wang (1999) and Warr (2002) identified industrial growth and service sector growth as having the greatest impact of poverty reduction in Asian developing economies. In spite of the vast and growing interest of poverty reducing potential of agriculture development in literature, the Africa region has witnessed limited attention. This present re-examination analyses the conceptual linkages from agricultural productivity to inclusive growth in Africa, and also, an empirical impact of agricultural productivity on rural, urban and dollar poverty in Africa.
The remaining part of the study is structured as follows: chapter two addresses the theoretical linkages and association with existing literature on poverty-agricultural productivity nexus. The third chapter outlines the conceptual framework reflecting channels and the processes through which agricultural productivity translates into poverty reduction and inclusiveness. Chapter four of the study addresses the methodology where the relevant empirical model was adopted to validate the thrust of the study. The fifth chapter comprises the discussion of empirical results and relation to extant theories and studies. The last chapter, six, concludes the paper with relevant policy recommendation rising from the empirical results. Using data from a nationwide Nepal Living Standard Survey 2004, they first estimate householdspecific productivity per worker under both Cobb-Douglas and translog production functions.
Second, the paper identifies the determinants of productivity. Third, they explore a theoretical link between productivity and poverty using Sen's poverty index and find empirically that productivity growth substantially helps poverty reduction.
Finally, the integrated effects of changes in productivity determinants are found to be stronger than the outcomes of sectorial policies taken in isolation. Also in reviewing the relevance of food insecurity, and environmental degradation are also analyzed. The challenges that must be addressed including, how best to intensify agricultural production, the types of technologies to promote and the imperatives of production efficiency and intra-regional trade were examined. Testing the impact of deforestation on aggregate agricultural productivity, Ehui and Hertel (1992), used an aggregate data from Côte d'Ivoire (the country with the highest annual rate of deforestation in percentage terms, i.e. 6.5%), average yield function is estimated which permits a variety of specific hypotheses to be tested. Results indicate that deforestation in the current period contributes positively to yields, and that increases in the cumulative amount of deforested lands cause yields to fall. This aggregate evidence confirms soil scientists' findings that crop yields increase after slash and burn deforestation because of the nutrient content of the ash.
Yields decline over time because of the removal of organic matter, erosion, and movement of cropping activity onto marginal lands. Aggregate yields were also found to respond positively to fertilizer applications, but with diminishing marginal productivity. Computed elasticities show that yield response to cumulative deforested land is quite 'elastic'. Other things being equal, a 10% increase in cumulative deforested land results in a 26.9% decline in aggregate yields.

Conceptual Framework
This section illustrates the three key roles agriculture can play in lowering poverty and promoting growth inclusiveness. These channels include: stimulating economic empowerment, creation of markets and fostering sustenance. In addition, the study discusses analytical tools that can help African economies examine these links and determine how agriculture can be leveraged to achieve more inclusiveness in the region.
There is a strong relationship between agricultural stagnation and poverty in sub-Saharan Africa. products. Growth of agriculture, of agricultural production, and of agricultural incomes helps the rural poor, and hence alleviates poverty. It also helps the non-poor, in some cases more than the poor (see figure 1).
Agriculture is said to play a key role in promoting inclusive growth. This is achieved primarily by stimulating economic growth, reducing poverty, and creating employment for millions of people in developing economies. However, its potential for future poverty reduction through these transmission mechanisms depends on the extent to which agricultural productivity can be increased where it is most needed.
Productivity growth can catalyze a wide range of direct and indirect effects that mediate the   Our estimation procedure began with examining the strength, pattern and direction of collinear relationship among the explanatory variables. The study attempted this by conducting the pairwise correlation matrix in table 1. The result shows no serious problem of collinear relationship which implies that our model is void of multicollinearity and the specific influence of our regressors is distinguishable. Likewise, the pairwise correlation matrix provides an insight on the likely impact of agricultural productivity on the indicators of inclusive growth. The preliminary evidence shows a negative correlation between unemployment and food production index; similar evidence also obtained between agricultural exports and unemployment. This implies that increased food production index and agricultural exports promote inclusiveness in African economies. Similarly, the variance inflation factor analysis was adopted to corroborate the results obtained using the pairwise correlation matrix; as the former provides a standard rule of examining the extent of collinearity among the exogenous variables . In order to ensure no serious problem of multicollinearity occurs, the variance inflation factor must be less than five and the degree of tolerance greater than 10 percent. An examination of the result presented in table 2 shows our explanatory variables do not exhibit any near or exact collinear relationship. Table 3 presents the regression analysis results for the static panel data model using the fixed and random effect specification. The choice of either fixed or random effect specification for the model specified in the study was based on the Hausman test for model reliability. There are basically five sets of results reflecting the five indicators of inclusiveness. Evidences from the result indicates that food production index and per capita income were significant in reducing both rural and urban poverty in Africa. In all category of poverty indicators assessed, food production index has the largest effect on urban poverty while per capita income posed same effect on rural poverty. Also, domestic credit impacts significantly other indicators of inclusiveness except urban poverty. A 1% increase in domestic credit reduces rural poverty, poverty headcount ratio $1.90 a day and poverty headcount at national poverty lines by 6.6%, 8.3% and 5.8% respectively. Similarly, a 1% increase in food production index reduces unemployment by 1.79%, rural poverty by 3.54%, urban poverty by 5.28%, poverty headcount ration at $1.90 a day by 8.11% and poverty headcount at national poverty lines by 4.62%. On the other hand, agricultural raw material exports, though significant but increases poverty incidence in Africa for all categories of poverty indicators. This will not be unconnected to the weak supply response and income inelasticities associated with export of agricultural commodities (Ogundipe, 2016;Cervantes and Brooks, 2009). In addition, the result indicates that institutions in Africa played a minimal role in reducing poverty, as it could not significantly lessen the incidence of poverty below $1.90 a day. This results needs to be taken with caution, as $1.90 a day underestimates the extent of absolute poverty in African economies. It was apparently revealed that institutions could meaningfully impact only one indicator of poverty-poverty headcount ratio at $1.90 a day. This indicates the weak level of institutions in most African economies in ensuring efficient income distribution and inclusiveness in Africa.  Table 4 presents the regression analysis results for dynamic panel data model using the system-GMM approach. This is necessitated to overcome the problem of endogeneity of the income and institution variables in the model, thus making the outcome more preferable to the static approach. Similar to the foregoing analysis, the system GMM specification was attempted for the five models specified which is reflective of the five measures of inclusiveness adopted. The  suggesting that growth in per capita (economy-wide) is itself driven by growth in agricultural sector income. Since agriculture is heavily subsistence in Africa, it hence implies that growth in per capita income is generated by other sectors or such agricultural growth does not occur in the space of the poor. This is reflective of the level of income inequality in African economies; national income growth impacts only the livelihood of urban population. It hence suggests the non-shared contribution of the rural agrarian population to national economic output and receives no shared benefits. Finally institutions were found to be a significant factor in reducing rural poverty and national poverty headcounts. In order to ensure the robustness of our parameter estimates, the study adopted some specification diagnosis tests, these includes the Arrelano-Bond test for autocorrelation, test of instruments validity and the F-test for the overall significance of our regressors (Akinyemi, Osabuohien, Alege and Ogundipe, 2016). The Arrelano-Bond test is conducted on the differenced residuals in order to purge the unobserved and the perfectly autocorrelated idiosyncratic errors. This is shown as AR(1) and AR(2) at the lower panel of Table 4, the significance of AR(1), and not necessary AR (2), implies that the successive values of the residuals are not serially correlated. The Sargan and Hansen J tests assess the over-identifying restriction of whether our instrument vector is exogenous, the test statistics failed to reject the null hypotheses, hence, the validity of our instruments is guaranteed. Finally, the F-statistic, a small sample counterpart of the Wald (Chi-Square) statistics shows that the exogenous variables jointly explained significantly the observed variation in energy security in Africa.

Conclusion and Recommendation
This study examined the effect of agricultural productivity on poverty reduction and inclusiveness in Africa using the dynamic panel data approach. The System-GMM estimation technique was adopted and preferred to the traditional OLS pooled regression and the static panel approach with the view of resolving the endogeneity problem inherent in the specified model.
Specifically, agricultural productivity was captured using agricultural value added per worker and food production index. The former suggests the relative contribution of farmers or agriculture dependent population to national growth while the latter covers the production of food crops that are considered edible and that contain nutrients. It simply suggests the availability of food for sustenance. On the other hand, poverty and inclusiveness was captured using five indicators: rural poverty, urban poverty, dollar poverty, unemployment and national poverty. The conceptual framework of the study identified three main linkages via which agricultural productivity leads to poverty reduction. This include: i. income empowerment, ii.
Market expansion, and iii. Sustainance enhancement. This implies that developmental efforts focused at enhancing productivity of livelihood and spaces of the rural poor results in increase of rural income, elimination of transgenerational poverty, equitable access to social and economic services., This will also bring about enhanced forward and backward agricultural linkages, expansion of non-farm sector income, food affordability, reduction in social discontent and conflicts, reduction in undernourishment and severe waste in adult and children.
Available evidences from the empirical investigation indicates that agricultural value added per worker contributes significantly to reducing unemployment and rural poverty in Africa. Since poverty is firmly entrenched in the rural areas and agriculture constitute the main income source for the 1.4 billion extremely poor people., an enhancement of the productivity of agriculture literarily regresses the incidence of poverty. This is consistent with Christiaeusen and Demery (2007), Bresciani and Valdes (2007) and Montalvo and Ravallion (2009) but in contrast with Ravallion and Datt (2002). On the other hand, food production index does not yield a significant reduction in rural poverty, though, this was obtained in other indicators of poverty. This result implies that food accessibility and affordability will pose a major threat to the non-farm poor, as larger proportion (if not all) of their income is spent on consumption.
Moreover, GDP per capita was found to significantly reduce urban poverty and dollar poverty.
This reflects the height of income inequality and non-inclusiveness in Africa. Since agriculture is predominantly subsistence and growth in one sectors are not easily transferred to another due to market segmentation and geographical remoteness; rural farmers can hardly share in economic growth. In the same manner, domestic credit to private sectors and institutions were significant in reducing all categories of poverty. It implies that development programmes targeted at enhancing agricultural productivity should encompass strategies for accessing credit in order to boost the asset base of rural farmer for a large scale commercial production. Also, appropriate macroeconomic policies and institutional framework quality needs to be put in place in order to boost provision of social services, equitable land and credit access.