Can We Trust U.S. Economic Statistics?
Grigory Khanin, Igor Dobrovolskiy
In the international academic community, U.S. economic statistics enjoy a reputation for exceptional reliability, with the United States consistently placed among the highest-ranked countries for accuracy in measuring GDP dynamics. To evaluate the foundation of this confi dence, the authors draw on their experience conducting alternative GDP estimates for the USSR, Russia, India, and China. This study examines the precision of U.S. economic statistics over the period 1970–1987 in three sectors – industry, construction, and agriculture. GDP volumes at current prices were defl ated using the manufacturing output price index for the processing industries and the fuel-price index for the extractive industries. The resulting alternative estimates of industrial GDP were compared with the offi cial fi gures published by U.S. statistical agencies. The offi cial GDP index in construction was checked based on the dynamics of housing commissioning and the share of housing construction in the total construction volume. In general, the analysis of statistics on the dynamics of US economic GDP confi rmed a high assessment of its reliability for the period 1970–1987, with the exception of price statistics.
Doubts about the reliability of American statistics for 1970–1987 arise when comparing favorable offi cial indicators of GDP dynamics with a large increase in the US national debt and a small (3-4%) share of construction in GDP for the period from 1990 to 2010. To assess the real dynamics of US GDP, natural indicators of production were adopted. They allow us to move on to calculating the dynamics of production in sectors of the economy that raise doubts about their reliability: manufacturing and construction. The greatest diffi culties arose with assessing the dynamics of the service sector. The authors assumed that labor productivity in it remained unchanged.
The calculations of the dynamics of production and labor productivity for 1990-2010 by sectors of the real economy showed an unprecedented drop in labor productivity in American economic history, which indicates deep troubles in the American economy during this period. The calculation of an alternative estimate of GDP dynamics revealed minimal growth indices: by 8% in the 1990s and by 4% in the 2000s, instead of the offi cial growth of 38% in the 1990s and 18% in the 2000s. Both estimates reveal a signifi cant slowdown in economic development in the 2000s. Possible reasons for this are considered. A conclusion is made about the most serious defects of American economic statistics in the period from 1990 to 2010.