Citizens do not usually have accurate information on population income distributions and their income rank position, and hence create beliefs of those measures based on partial information. The literature of income comparison has found that reference income of co-workers and personal network have a significant impact on those perceptions, but other influential factors remain unknown. Are tax schedules used by taxpayers as signals on the actual income distribution? Under certain assumptions on how the government designs tax system, changes on the tax burden applied to different levels of income may be used as signals of the actual income densities. Using an online experiment where workers face different tax systems, I aim to identify the potential impact of the structure of tax scales on perceived income distributions and self income rank.