i.e. evidence reflected in the components of income after the earnings manipulation.
As explained in more detail later, ceteris paribus, we expect the manipulated
components of income to be higher for firms with slightly positive earnings
as compared to firms with slightly negative earnings.