[gallery]
Price premiums for quality are common in specialty green coffee buying. But which effect has a cupping score differential of 2 points on prices? Would this premium change by country of origin and does it depend on the number of bags available of this specific lot?
If somebody asked me these questions, I would answer them based on my experience. However, with experiences come feelings. Probably feelings are not rational to discuss price implications of a billion dollar market. This is why I believe it is time for thorough analyses on the quality premiums in specialty coffee buying.
Wilson and Wilson provide such an analysis in “The economics of quality in the specialty coffee industry: insights from the Cup of Excellence auction programs” [2014, Agricultural Economics]. I name their SCAA lecture during which they discussed their paper, to be my edition’s favorite. While other green coffee buying lectures only mentioned the specialty premium the authors provided explanations for it.
Wilson and Wilson conclude on positive relationships between prices and altitude and prices and scarcity (number of bags of the lot). Moreover, they confirm a doubt I have carried with me, there is no relationship found between certifications, like Rainforest Alliance, and the specialty premiums paid during Cup of Excellence auctions! Or in other words: once we talk quality, we don’t talk certifications. Most importantly, the authors were able to estimate the effect of an increase in quality by one point on the price to be almost 16%!
Side note for all the readers who made it until here and do enjoy statistical modelling: I would like to see two robustness checks:
On one hand it is common knowledge that coffee prices are a function of Brazilian weather conditions (biggest producing country). Likely the variable “specialty prices” depends on the Brazilian weather as well – otherwise they could be lower than the C-Market. On the other hand the explanatory variable “number of bags” is a function of weather as well. How do estimates change once an index for drought in Brazil is included?