The paper addresses some difficulties in designing empirical research aiming to obtain an evidence of a real existence of Fiscal Illusions – the concept well known within a theoretical economic literature for more than a hundred years. There is an extensive literature trying to demonstrate empirically the presence (or absence) of fiscal illusion in the real world. The results are not convicting, according to a classical Oates study (1988) and it seems to be confirmed by our review of current papers dealing with this topic. The contemporary empirical literature includes numerous highly relevant, instructive, and comprehensive papers dealing with many facets of fiscal illusions. One aspect, however, seems to be omitted. Econometric studies comparing within sample of governmental institutions on the same level certain measures as explanatory variables (e.g. complexity of the tax system, the share of subsidies, etc.) with the amount of public budgets, and the dynamics of their growth are mainly used in order to indicate the fiscal illusion. In fact, fiscal illusions may be used/initiated not as an attempt to increase (annual) budget expenditures but rather because of a) avoidance of voters resistance against given taxation volume; b) to govern easier, with lover transaction costs (e.g. with less pressure to demonstrate effective use of public funds…). These measures are inherently difficult to quantify. We discuss possibilities to replace their direct measurement by using some auxiliary indicators, such as number of officials and/or public employees, the relative level of wages in the public sector, the relative proportion of certain types of expenditure, etc.