This is a clearly written and nicely organized upper-division macroeconomics textbook. Mankiw uses plain English and simple math to model the macroeconomy in the short-run (the IS/LM model), the long-run (the AS/AD model), and the very long run (the Solow growth model). He also devotes a lot of space to the Mundell-Fleming model of international trade and finance. One of the best features is the frequent use of short case studies that apply economic theory to "real world" problems such as the Great Depression or the Japanese slump of the 1990s. Mankiw's views are mainstream -- he doesn't even hint at the existence of alternatives such as Austrian economics or neo-Marxism -- but he is non-dogmatic about policy and quite candid about the limits of what economists really know about the economy. His book is a small masterpiece of clear economic writing for undergraduates.
So why did I give it only four stars? I was disappointed by the relative neglect -- in spite of the many "case studies" -- of the micro-economic, historical, and institutional realities that underlay the graphs and algebra of conventional macroeconomic analysis. Let me give two examples of what I mean:
-- According to Ben Bernanke, Asian countries responded to the financial turbulence of the 1990s by amassing huge foreign exchange reserves to defend their currencies against future attacks. These savings have for the most part been invested in the U.S., where they have financed trade deficits and fueled asset bubbles in the equities and housing markets. In other words, capital has flowed from relatively capital-poor countries to a capital-rich country, where it has paid for consumption binges.
-- According to Robert Pollin, two decades of union-bashing, downsizing and free trade have led to widespread job insecurity in the U.S. With workers too intimidated and too worried about jobs to press for wage increases, the economy was able to grow in the 1990s without triggering a round of inflation, and the benefits of this growth were skewed towards upper-income groups. In other words, extra-market power relationships in the workplace directly affected macroeconomic performance and income distribution.
I don't doubt that these developments can be captured and analyzed in the IS/LM or AS/AD framework. I'm not so sure, however, that many people steeped in this mode of analysis would have expected these developments ex ante. That would have required a knowledge of history, policy responses, and specific markets that is difficult to capture in abstract models. For my taste, any approach to economics that focuses on algebraic relationships between economic aggregates to the semi-exclusion of history and institutions is just too "otherworldly" to be satisfying. But maybe that's a problem with the way my mind works, not with macroeconomics. It certainly doesn't mean that Mankiw's book is anything less than excellent. Any student interested in learning basic macroeconomic analysis should read it.