The article argues that sociological analysis of justice and fairness perception phenomena requires a closer examination of their purported relationships with everyday descriptive knowledge available to social actors. The results of an experiment studying relationships between individual judgments on fair division of putative centralized transfer payments addressed to different groups of the general population and subjects’ ordinary pre-existing knowledge of relevant social and economic facts are presented and analyzed. The experimental results obtained indicated that in most cases no significant correlations between individual judgments on just distribution of money transfers aimed at 1) equalizing average salaries for those working in different economic sectors and 2) Internet-access levels for different age cohorts, on the one hand, and relevant everyday predictions for salaries and Internet-access distributions, on the other hand, were observed. The weak-to-moderate correlations between individuals’ everyday cognitions and their justice-related judgments were revealed only in few cases for particular groups of hypothesized money transfers beneficiaries and could have a direction contrary to theoretically expected one. The results obtained give some indirect support to a “cognitivist” theory of moral feelings (R. Boudon) and cast doubts on popular broad interpretations of “the Thomas theorem”.