Credence goods, efficient labelling policies, and regulatory enforcement

We analyse the optimality of information revelation of hidden attributes of "credence goods" via alternative labelling procedures. When consumers are heterogeneous in their willingness to pay for the hidden attribute, producers can either self-label their products, or have them certified by a third party. The government can impose self or third party labelling requirements on either the "green" or the "brown" producers. Our benchmark model develops a condition that links the optimal imposition of third party labelling to the relative market share of each type of the good under complete information. We extend our analysis to incorporate asymmetric information and cheating by the producers. When corrupt producers can affix spurious labels, the government needs to supplement the labelling policy with costly monitoring activities. We find that mandatory self-labelling schemes generally dominate mandatory third party labelling, unless the "market share effect" greatly exceeds the "incentive-to-cheat effect".