In fact, current studies show that people who say they read food labels also say that they have better food habits. However, I want to emphasize the word “say”. They say that they are concerned with their food and that they read the labels, but when they draw up a list of what they eat, we see that there is a bias toward desirability. People often tend to make things appear rosier than they really are. We have not conducted any observational studies. Currently, direct observations of the effect of a claim or a label on behaviour is something that is not found in the literature. If I recall correctly, one study showed that it did not necessarily have a significant impact. This was a specific case where a label was associated with a food item. When people did not know that the nutritional value was being measured, the label was ineffective. This is the important point. However, when people know that things will be measured, there is a bias towards to the measurement.
On March 1st, 2011. See this statement in context.