昨日エントリでは実質所得の伸びの過小評価を指摘した論文を紹介したが、フェルドシュタイン御大もその問題を扱った表題のNBER論文(原題は「Underestimating the Real Growth of GDP, Personal Income and Productivity」)を上げている。ただし、昨日紹介した論文では実質化に使うデフレータの問題を取り上げていたが、ここでは品質向上や新製品導入の計測問題を取り上げている。以下はその要旨。

The problems involved in estimating real output that I discuss in this paper cause the official government statistics to underestimate of the rates of growth of real GDP, real personal income, and productivity. That underestimation is important not just to economists trying to understand where the economy is going but also to the broader public and to the political system.
The understatement of real growth reflects the enormous difficulty of dealing with quality change and the even greater difficulty of measuring the value created by the introduction of new goods and services. Despite the vast amount of attention that has been devoted to this subject in the economic literature and by the government agencies, there remains insufficient understanding of just how imperfect the official estimates actually are. It is important for economists to recognize the limits of our knowledge and to adjust public statements and policies to what we can know.