No, you did not guess that people would use aggregate income tax stats among other data in estimating the income of the superrich. I would not have pointed out that you were uninformed had you said anything like that. 1) Roads are not the only government infrastructure provided by government 2) Roads and auxiliary government infrastructure are necessary for horse and camel transport of goods in agricultural societies, and other (lesser, but still critical) government infrastructure of various kinds is necessary for horse or camel transport of goods to established markets in non-agricultural societies (pastoral, nomadic, etc). Ferries, bridges, stables and rest areas, bandit security, communication, medium of exchange and methods of credit, management of common grazing areas (Hardin's Tragedy, specifically), - - it's a long list. No governments, no markets. In small communal societies without markets, there is personal and community accountability without government. So? And why do you replace "government" with "state" when replying to posts? So? Are you trying to present an argument based on assuming the analysis in that book does not recognize tax evasion as a factor in estimating incomes? You will learn a lot about that idea, if you read the book. Among the things you will learn is that "redistribution" is not an accurate term - squelching the growth of inequality is not a matter of redistribution. It is not what you did guess. What you did guess was that rich people's individual tax returns were the only source of info, and that the entries on them were accepted without analysis or correction. You didn't have the researchers "making use" of tax records among a plethora or other data, but naively accepting them as sole and unquestioned authority without prudent analysis.