Read: Economics - The User's Guide
Chang's Economics: The User's Guide is less a guide to economics as it is practiced but more to an economics that should be: A pluralistic study of the economy, not one single approach to analyze everything.
Chang introduces to some of the non-mainstream approaches to economics and shows where they differ from the mainstream by discussing concrete current economic issues (with a slight emphasis on macroeconomics) and how we got there. As such, the Guide is rather a work in economic history and comparative economics. The presentation of relevant statistics on a wide range of countries to provide context is commendable.
Underlying the obvious narrative, an introduction to economics, there seem to be two theses. First, economics is inseparable from politics. Politics provides the institutional context, the rules of the game, economic analysis has to be performed within this frame. Economics, however, provides the reasoning for many political decisions, it “is a political argument.” The two fields cannot be without the other. Second, as a result, all economics is normative. There is no value-free economic analysis. “[T]here are no objective truths in economics that can be established independently of political, and frequently moral, judgements.” Hence, “[e]conomics is not - and can never be - a science.”
While I mostly agree with the first thesis, I do not agree with the second one. Economic analysis (in particular of microeconomic issues) can be purely descriptive, positive. Without some understanding of the causal links, the mechanics of the economy, the decision-making behavior and process a meaningful normative analysis would not be possible. I would also argue, that, e.g., experimental economist are not just going through the motions of the scientific process of knowledge generation, they are scientists: Observing, hypothesis generating, falsifying, and theory building scientist.