Jin and Leslie (2003) analyzes the impact of the introduction of government issued hygiene grade cards on restaurant cleanliness in Los Angeles. They find that cleanliness did in fact improve significantly after the government started using these grade cards.Restaurant revenue is significantly affected by the posted scores,with restaurants that received an “A” grade getting a. percent increase in revenue from mandatory disclosure, while restaurants that received a “B” got only a . percent increase, and restaurants that received a “C” got a 1 percent decrease. This provides evidence that consumers do pay attention to this additional source of information, and that market demand was changing as a result. Jin and Leslie (2009) extends this study by considering whether reputational effects alone are adequate to explain this restaurant behavior. The paper assumes that restaurants affiliated with chains have an incentive to free ride on the chain’s reputation. Thus, franchised stores tend to have lower hygiene scores than company-owned stores. After the introduction of the grade cards, the authors assume that such incentives would go away, because consumers are able to infer hygiene directly from the posted scores. From the data, the authors find evidence that the introduction of grade cards causes hygiene scores to increase more for non-chain stores than for chain stores, lending support to the idea that chain stores care more about reputation in the absence of grading cards. And in comparing the effect of reputation versus the effect of the grade cards, the authors find that hygiene improvements due to reputation were 70%as large as hygiene improvements due to the posted grade cards. The significant changes in hygiene after grade cards were introduced thus shows that certification can send a clearer output to the market than mere reputational effects, and that reputation alone may not be sufficient to solve adverse selection problems. Our model echoes this result, by showing that certification can increase social welfare even if the market can learn via the agent’s reputation.