Resiliency - Building Coffee Capacity At Origin: February 2018

Feb.27, 2018

Coffee cherry in Rwanda

What does a 240 cherry floor price mean in Rwanda? More poverty for coffee farmers. NAEB has announced a 240 RWF/KG cherry floor price (farmgate price) for coffee cherry for the 2018 season despite research showing that around 25%, one fourth, of Rwanda's coffee farmers will be losing money on every coffee cherry they sell. (See Figure 1 below.) When viewing this figure, keep in mind that the threshold used for "profit" here is any amount above 0. The data is credible, as it is based on a 1024 sample survey (AGLCa, 2016), which included estimating cost of production for coffee at 177 RWF/KG cherry (AGLCb, 2016). Most businesses, including the exporters who are pressuring NAEB to keep the cherry price low, would go out of business if profit levels fall below 20%. The red bars in the Figure below would be much higher if we assumed farmers had the right to have 20-30% margins. Then Figure 1 would show that nearly half of Rwanda's farmers are "in the red" at a floor price of 240.

Figure 1: Percent of Coffee Farmers Making Above 0 Profit at Various Cherry Prices

So the 2018 season begins with these "barely getting by" prices for farmers, forced to be this low by exporters who will gorge themselves on 50% margins. Meanwhile, the exporters will publicize their exciting 'investments' in water projects to improve waste water treatment at one coffee washing station, and 'investments' in a training project that brings cows to smallholders. These 'investments' will be so exciting they will end up in articles on the covers of glossy coffee trade press magazines and even in Harvard Business Review case studies about "social enterprises" doing the right thing for poor coffee farmers in Rwanda.

Please don't be fooled. The powerful in Rwanda are hungry enough to eat every coffee farmers' lunch. They would rather invest in projects on the fringe, than work to improve their marketing and sell more specialty coffee at higher prices. They cannot see their way to paying a few cents more, just $.19/lb. green coffee, to ensure that farmers are motivated to plant seedlings, prune trees, conserve their water and buy a cow if they would like to.

Please don't be fooled by projects and trainings. 57% of Rwanda's coffee trees are owned by commercially oriented farmers with .5 hectares or more planted with coffee. Many of them could increase their yields by two or three times this season -- if it is commercially viable to do so. They know how. But they choose not to because with cost of production at 177/KG cherry, 240 RWF/KG cherry means 64 RWF/KG cherry net income, or a 26% gross margin and they have better options in banana, livestock and other farming businesses. Many will use low-input strategies or neglect the coffee trees completely because they can afford to transition to other business opportunities.

The smallest farmers, with an average of 180 - 200 trees, will not be so lucky. They will suffer yet another year of losses for all their effort on their coffee trees. For them, a training may minimize their costs per unit of cherry produced and therefore minimize their losses. But leaving the coffee and using the same amount of time to pick potatoes as a day-laborer would probably bring more cash into the household.

You might ask "why?" Why do companies implement projects and trainings, but refuse to pay $.07/KG cherry more, (equal to +$.42/KG green coffee, or +$.19/lb green coffee) for a supply of quality, fully-washed coffee? For some it is greed to make a few pennies more this year that drives such decisions. For other exporters it is short-sightedness. Despite the research (AGLCc, 2016) they do not understand that their insistence on cheap cherry means death of a vibrant option for increases in yields and supplies of high-quality cherry, plus rural development in Rwanda. Meanwhile, they keep their options open to move over to Ethiopia or Tanzania once Rwanda's commercially viable coffee farmers have all pulled out their trees.

Pennies -- they are unwilling to pay pennies of investment in the coffee farmers via the coffee cherry price, (i.e. the farmgate price).

Specialty buyers of Rwandan coffee should be outraged at the downward spiral the 240 price can precipitate: lower prices, leading to lower yields, leading to lower motivation from commercially motivated Rwandans living in rural areas. Rwanda was poised to reverse this vicious spiral in the 2017 season. At this time last year NAEB pegged the floor price first at 270, a promising start. But then they let it decline to 240 later in the season. The weighted average for 2017 was 249. Starting this 2018 season at 240 starts to look like we are headed back to the bad 'ol days of 2015 and 2016. In those years the floor price was 170 and 150 respectively.

If you buy Rwandan coffee, you know the FOB price, (called the export price in our scenario illustrated below, $4.30/KG green). Please, demand to know the farm-gate price from your supplier. Farmers want 300 RWF/ KG cherry. Research (AGLCa, 2016) shows 300 RWF provides sufficient motivation for farmers to invest in best practices, which results in high-yielding plants and more specialty grade coffee. A win-win for all. Any exporter paying less than 300 RWF for coffee bound for the specialty market should explain to you why coffee farmers deserve margins only half the size of exporters (Chart 1 vs. Chart 2 below).

Chart 1: 240 RWF/KG Cherry

Chart 2: 300 RWF/KG Cherry

Feb. 20, 2018
Great blog today, "The Toyota Way or Entropy?" by David McKenzie, an economist at the World Bank, on the World Bank Development Impact blog. McKenzie takes the most typical criticism of any training, which is that it won't stick, and tests it. Specifically, McKenzie and co-authors Nicholas Bloom, Benn Eifert, Aprajit Mahajan, and John Roberts test whether an intensive management intervention in Indian textile weaving plants can be shown to have staying power. The simplified, contrasting possible outcomes are the“Toyota way” , with systems in place for measuring and monitoring operations and a continuous cycle of improvement vs. the alternative of entropy, or a gradual decline back into disorder. "One estimate by prominent consulting firm Boston Consulting Group, is that two-thirds of transformation initiatives ultimately fail." Fortunately, McKenzie et. al. find different results.

The study results shared in their new paper are somewhere in between the two outcomes described above. There is some entropy, but some improvements are "sticky", and even 8 years later are still in practice. In short, there was more persistence in management practices than the researchers anticipated. We (at Artisan) were most interested to see which practices were the ones that had staying power. They relate very closely to improvements in quality and inventory levels:

recording quality defects in a systematic manner (defect-wise);
having a system for monitoring and disposing old stock; and
carrying out preventative maintenance.

We're excited to see this list, as #1 and #3 have been key success areas in the Lean at Origin two-year pilot project carried out at washing stations in Rwanda and Burundi.
#1 - recording quality defects in a systematic manner: In a washing station in Rwanda, the first "big KAIZEN" a work team addressed was to devise a way to measure the defects coming out of the depulping machine. At the very least, monitoring the level of defects alerts the machinist as to when he needs to adjust the discs. In an advanced scenario, the level of defects after the depulper can be an indicator of the overall quality of the cherry the washing station is receiving. This has important implications for farmer education and washing station profits to the extent that revenue is dependent on quality.

#3 preventative maintenance: At a dry mill in Burundi, the operations manager recognized right away how impactful a program of regular maintenance could be for the depulping machines at the 30-some washing station delivering parchment to their mill. No doubt, he also recognized how the quality of the parchment from these 30 washing stations would be improved if machine down-time became less of an issue.

#1 - photos of recording quality defects in a systematic manner at KOPAKAMA: