I learned about value chain analysis and the whole business of “making markets work for the poor” in my first job in Bangladesh. I think I also developed an understanding about how hard it is to make market work for the poor. Probably I shouldn’t have started in such a pessimistic tone. But I feel it is better to be aware of the limitation of a tool before starting to use it. And again, I don’t want to say this approach cannot be used as a means of poverty reduction. But we need to balance between market and other approaches whenever necessary.
Nevertheless, value chain analysis is certainly an instructive tool to pinpoint the key problems in different parts of the supply chain that prevent poor people to benefit. It eventually helps to design targeted actions to fix those specific issues. My assignment here in Mongu is to participate in a research on the Barotse Floodplain fish value chain.
When we arrived in Mongu, the first round of data collection was already complete. The questionnaire designed to collect different value chain information including fisher/trader profile, fish species, products, prices, costs, trade channels and the enabling environment was very long. Besides, the informal nature of the business and varied measures and methods of transactions that are in place presented several challenges to collecting the correct information, particularly in the area of price and quantity.
Fishers and traders use different measures of volumes to sell their products. They do not use weight. So there is an element of inaccuracy built-in as fish size, shape and dryness will determine exactly how much fish goes in one container. Again, the standard bowls used for selling fish can be just full or cramped and heaped. Different types of containers are used to measure different types of fish. How do we get consistent measure? To further complicate the problem, there are several species of fish. The price of fish varies by the species and by size of the fish within the same species. Price also depends on the channel that is used for selling. How many options can we include without frustrating the data collectors and the respondents? All these complicated issues rendered some of the first round data not so useful.
However, in the analysis workshop in which all the data collectors participated, we had a long discussion on how to standardize the measure and how to deal with the species-size-measure complexity to get the closest estimate. Based on the discussion, we designed the questionnaire for the second round, which we are hoping to pretest in next week. I am praying that the questionnaire will pass! We also need to find out the best way to input the enormous data set. Excel is not suitable for inputting such a large amount of data in the same row. I am planning to use Formhub for easier input. But I found some issues with data output from Formhub. I am thinking about consulting with Modi Research group for some advice.
My learning from first few weeks is that designing a questionnaire may seem a straightforward task, but actually it is not easy, particularly if we need to gather compex market information. A great deal of effort and time needs to be invested to devise a questionnaire that works!