Accounting for non-market activities in the distribution of income: An empirical investigation☆