Consumer Demand for Labor Standards, Part III 

  Michael Hiscox and Nicholas Smyth , guest bloggers 

 Continuing our discussion begun yesterday and the day before on labor standards labeling, perhaps the most important comments we received at the workshop had to do with how we might design our next set of experiments. It is very difficult to do anything fancy when it comes to in-store experiments. It could never be practical (and ABC would never give permission) for us to randomize treatments to individual items or brands on, say, a daily basis. The manner in which products are displayed (grouped by brand), the costs associated with altering labels and prices, and the potential problems for sales staff (and the potential complaints from frequent customers) impose severe constraints. Several workshop participants suggested that we conduct the next set of experiments through an online retailer. That way we might be able to randomly assign labels (and prices) to customers when they view product information on a website and decide whether or not to make a purchase. There would still be plenty of difficulties to iron out, as was quickly noted (e.g., making allowances for customers who attempt to return to the same product page at a later point in time, and for customers who "comparative shop" for products at multiple retailers). But this seems like the way to proceed in the future. 

 On a related theme, we noted that Ezaria.com, an online retailer run by Harvard students, is already planning to track a variety of economic data on its customers. Ezaria has a mission which involves providing markets for independent artisans from the developing world and donating 25% of profits to charity. At a minimum, looking at data on whether a customer is more likely to make a purchase after being shown the company's "mission" page (that explains their policies) would provide some measure of consumer demand for companies that source from high-standard producers. Perhaps we can persuade Ezaria to cooperate with us in a future experimental project. Or perhaps we can arrange the experiment with an even larger online retailer, with customers who are not so obviously self-selected as socially conscious.