Consequences of the Five-a-Day Campaign: Evidence from French Panel Data

'5-a-day' public information campaigns have been implemented in many countries to promote the consumption of at least five portions of fruits and vegetables (F&Vs) per day per person as part of a healthy diet. This study proposes an ex-post evaluation of its impact on household purchases of F&Vs for food-athome consumption, using household scanner panel data and a before-after identification framework. It uses a dynamic tobit panel data model to account for non-participation, habit formation and household unobserved heterogeneity. Over the 2002-2010 period, the portions of F&V available each day for consumption by a household member have increased from 2.4 to 3.3 (+0.9). Our estimates reveal that 43% of this increase is due to the 5-a-day campaign (+0.38 portions). The largest effect is observed for fresh fruits (+0.16 portions), processed vegetables (+0.08 portions) and natural fruit juices (+0.08 portions). As a negative nutritional consequence, there has also been a small increase in fruit drinks with added sugar (+0.05 portions).