---

Inefficiency and injustice of subsidy system in Iran over the years induced the government to implement targeted subsidy law since 2010. As a result, the share of poor households receiving government subsidies decreased. Utilizing the micro-data of Households’ Income and Expenditure Surveys (HIES) from 2005 to 2014 and applying seemingly unrelated regression(SUR) technique through feasible generalized least squares (FGLS) method, this study examines the effect of targeting subsidies on households’ consumption combination. The results show that the shares of necessary goods in households’ expenditure have increased. According to Engle theory, these indicate reductions in households’ welfare. In addition, the “subsidy” dummy variable has the most effect on food expenditure share among different goods groups. One reason for the welfare reduction is higher increase in relative prices compared to increase in income due to subsidy. Since, the permanent income is of the highest effect on allocation of households’ budget among different goods groups; therefore, policy makers should focus on increase in the households’ real permanent income instead of inflationary policies, which increase nominal cash subsidies.