In the recently-concluded conference of the Asian Society of Agricultural Economists (ASAE) held in Dhaka, Maros Ivanic and Will Martin (from World Bank) dealt with an interesting area of investigation on which researches seem to be rare. It is the relationship between growing agricultural productivity and poverty reduction. They argued that various components of total factor productivity improvements in developing countries bear important implications for global poverty reduction. Using a Global Computable General Equilibirum model with 315,000 households in 31 countries, the authors showed, in percentage points, the change in the poverty rate at $1.25 per day resulting from a one per cent increase in total factor productivity in agriculture, industry and services. Since the sample is reported to cover three-fourths of the world's poor, one could perhaps expect a reasonable approximation of the results in global context.
It may be mentioned here that an enquiry into the nexus between economic growth and poverty reduction is not new; it had long been an aspect of enquiry by economists. One could recall the remarkable researches done by Martin Ravallion and others to show that the impact of growth on poverty depends not on growth per se but on the sectoral source of growth. In other words, an impact of economic growth on poverty reduction is heavily influenced by the sector in which growth tends to take place. For example, growth that springs from high tech mining or manufacturing may not generate employment due to low elasticity of employment with respect to output. But in the hub of poor such as in agriculture or informal activities, growth could come to the help of the poor. Some other studies also point out that growth in rural areas may be associated with more reduction in poverty than the same happening in urban areas.
But all of those references relate to economic growth in a broader canvass. Maros and Will make a departure by focusing on total factor productivity changes across sectors and their implications in the realm of poverty reduction. Not many studies are in evidence in respect of this kind of relations. Changes in productivity - a la Maros and Will - would go to affect low-income households in three basic ways: (a) through changes in the productivity of factors that they employ in business, such as farms, that they operate; (b) through changes in the prices of goods and services that they consume, and (c) through changes in factor returns (say, wages) they receive for the factors for which they are net buyers or sellers. Another important point to note is that local productivity gains raise producer returns while more widely-adopted productivity increases mainly reduce poverty through reductions in consumer prices.
The key finding is that productivity gains in agriculture are generally, but not always, more effective in reducing global poverty than equivalent-sized productivity gains in other sectors. "This means that increasing GDP (gross domestic product) by identical amounts through increasing productivity in different sectors would lower poverty most if the productivity gain is in agriculture. Even when the size of the productivity gain in agriculture is not adjusted for the lower share of agriculture in majority of developing countries, poverty reductions through improvements in agriculture are on par with the gains from equally sized productivity gains in other sectors."
Interestingly, they find that benefits of productivity gains for poverty reduction in the countries that implement them are not very responsive to how many developing countries take part in productivity improvement. If a single country improves productivity, poverty falls in consequence of profit gains to producers. When more countries improve productivity, poverty falls at about the same rate. The outcome may sound surprising as reduced prices wipe out producers' profit gains, but bear in mind that such productivity gains benefit more consumers through lower cost of living.
By and large, according to the authors, an increase in productivity that increases GDP by one percentage point may have more favourable impacts on poverty rates if it arises in a sector that is more important as a source of income for the poor, or because of declining prices following the productivity gains disproportionately reduce living cost for the poor. It is clear that an increase in productivity gains has a bigger poverty reducing impact if it arises in agriculture than in other sectors.
Could we relate the results of the research by Maros and Will to the context of Bangladesh? We would argue that their results have important implications for Bangladesh in its attempt at reducing poverty. It is true that Bangladesh has substantially reduced poverty to account for 33 per cent now. Equally true is the fact that the share of agriculture to GDP is 20 per cent or less but the sector employs about half of the labour force implying pitifully low productivity. During the last few decades, the advent of modern technology in rice and ICTs (information and communications technology) have increased productivity through the channels cited by the authors. What is disconcerting, however, is the trend that productivity growth in agriculture has almost stagnated on account of various agro-climatic, socio-economic and other reasons. Unless the first generation technologies are replaced with a set of second generation technologies (such as Hybrid for HYV), the continuum of productivity gains might get a jolt. Agriculture in Bangladesh is still the occupation of the majority of the poor faced with twin traumas: small and marginal producers need to increase profits from marketed surplus on the one hand while the landless and urban poor need lower food prices on the other. An increase in total factor productivity in agriculture could possibly provide an answer to the dilemma by raising productivity and thus benefiting both groups. In fact, increase in total factor productivity in agriculture could kill two birds with one stone - helping both consumers and producers.
But as Bangladesh has been opening up quite rapidly with RMG (ready-made garment) and other export sectors contributing to a larger employment generation for the poor from rural areas as well as the growth of the non-farm sector as alternative pockets of employment, total factor productivity increase in those sectors hold no less importance. The message from Maros and Will is that increased productivity in agriculture is the key to poverty reduction.
Abdul Bayes is a Professor of Economics at Jahangirangar University.
abdulbayes@yahoo.com