Elephantstrunk

Owen Barder notices the FT confirming a comment on Tyler Cowen’s Fairtrade perplex. As “Tom” says in Tyler’s comments, Fairtrade suppliers are not perfect employers and buy coffee from neighbours when they can sell more than they can produce. So what?
The Fairtrade producers quoted are still better employers than other coffee producers. The problem coffee producers face is that they have less capital than almost all of the other players in the coffee value chain which means that they hold almost no bargaining power. The Fairtrade deal, guaranteeing a minimum price, is effectively an injection of capital.
Starbucks and others have exploited the demand inelasticity for coffee but its rise has done very little for coffee producers. Fairtrade is not an end in itself, schemes like Cup of Excellence and the culture that makes it viable offer a more secure future. Ideally it would end up as irrelevant as would Fairtrade wine.

The amount of hostility that such a market based solution can arouse leaves me non-plussed. Maybe economists don’t like the idea that the invisible hand works best when fuelled by enlightened self interest because it makes transactions harder to model. Opposition to food labelling requirements from economists might be similar. Whatever the reason, much of the commentary seems to get off on the wrong foot.

Firstly, even characterisation of Fairtrade coffee goes askew. Here’s Tyler Cowen from the post linked above:

In case you don’t know, fair trade sells a product at a premium price, under the promise that the workers are treated better and paid more.

He ought to have observed that consumers are in fact willing to pay more for fairtrade coffee but even that is not necessarily right. Sometimes fairtrade coffee is not even more expensive. A combination of the small cost impact of actually paying fairtrade rates and the marketing value of fairtrade as a differentiator could make it cheaper. The second part is wrong too. The promise is that the producer will be paid more for the coffee under most circumstances. The producers are also often workers and there are extra conditions on treatment of workers. The statement is not entirely wrong but the proposition is definitely about more than hand outs to workers.

Gordon Smith provides a good example of the next objection, most of the markup doesn’t go to the farmer. This seems wrong headed. If there is a large markup and most of the markup is not going to the farmer it suggests that supply of fairtrade coffee is too limited and this is the normal way markets satisfy imbalances. It also highlights how little the cost of coffee from the grower impacts the price of coffee to the consumer. The example of 2p from a 99p bar of chocolate going to the farmer is presumably meant to suggest that 2p is not much but to the farmer that could be a markup of 50%.

It might also be meant to suggest that the consumer is getting poor value but no other donation route is suggested. Charities are not that efficient in their transmission but its targetting where they will have trouble matching fairtrade. The idea should be to give the farmers the proverbial fishing rod while the objection is counting the number of fishes. Surely true believers in the market should see the discrepancy as an anomoly that will be ironed out by market forces.

The most interesting angle is Cowen’s idea that “Fairtrade” coffee makes non-fairtrade coffee effectively “exploitation coffee”:

How about a genre called “Exploitation Coffee”? You pay less, and they promise to treat the workers especially poorly. That wording is a less effective marketing ploy, but that is what quality differentiation and indeed “fair trade” boils down to.

It is well known that price discrimination can either raise or lower the average level of prices, but it does increase price dispersion. We can expect it to increase wage dispersion as well. It is harder to predict whether price discrimination will raise or lower wages at the bottom level of the scale.

By increasing output, fair trade can bid up wages for coffee producers. But fair trade also diverts some drinkers from Exploitation Coffee. If the switching effect is large, wages for producers of Exploitation Coffee can fall. Just as we have created two classes of market prices, so have we created two classes of market wages. If you believe that coffee producing firms have some degree of monopsony power, this is sustainable and again will increase profits but possibly worsen human misery for the poorest.

Later Cowen admits that this is a worry and without quantification is not an objection. While the fraction of coffee that is fairtrade is small I suspect the downside will be nearly non-existent. Even if the scale becomes significant, the focus on labour costs is misleading and needs a little more context. Coffee picking is not the only job, labour stipulations for fairtrade coffee, where they exist, usually specify no more than paying the minimum wage.

Two theoretical claims implicit in this criticism are testable. Simplest is the claim that retail coffee prices are relatively insensitive to producer prices. More contentious is the claim that as in poker, capital makes a difference to bargaining outcomes.