This is a well understood phenomenon in the world of Garbology [1] called the Lean-Cuisine Syndrome [2]
In one study garbologists found that people uniformly underreported the quantity of junk food they eat, and overreport the amount of fruit and diet soda they consume. Most people also underreport their consumption of alcohol by 40 to 60 percent; on the other hand, heads of households regularly exaggerate the amount of food their families consume. [3]
Or the well known phenomenon that when people are asked how many lifetime opposite-sex sex partners they've had, the average reported by men is about twice the average reported by women (though, by construction, they ought to be the same - just imagine a bipartite graph (undirected!) and count the lines going from left to right and compare with the number of lines going from right to left).
See eg. David Spiegelhalter, Sex by Numbers: What Statistics Can Tell Us About Sexual Behaviour, chapter 3.
See eg. David Spiegelhalter, Sex by Numbers: What Statistics Can Tell Us About Sexual Behaviour, chapter 3.