Integrating Expenditure and Income Data: What to do with the Statistical Discrepancy?

This discussion paper led to a publication in <A href="http://books.google.nl/books?hl=nl&lr=&id=rqM2GtTl9NcC&oi=fnd&pg=PA309&dq=allesintitel:+%22Integrating+Expenditure+and+Income+Data%22&ots=k7CbcJlBhn&sig=8cJoITJe8wPXfrHVXFl99x_biXQ#v=onepage&q&f=false">(D.W. Jorgenson, J.S. Landefeld, W.D. Nordhaus, eds.) 'A New Architecture for the U.S. National Accounts', NBER Studies in Income and Wealth</A>, vol. 66, 309-54, University of Chicago Press, 2006.<P>The purpose of this paper is to build consistent, integrated datasets to investigate whether various disaggregated data can shed light on the possible sources of the statistical discrepancy. Our strategy is first to use disaggregated data to estimate consistent sets of input-output models that sum to either GDP or GDI and compare the two in order to see where the discrepancy resides. We find a few “problem” industries that appear to explain most of the statistical discrepancy. Second, we explore what combination of the expenditure data and the income data seem to produce the most sensible data according to a few economic criteria. A mixture of data that do not aggregate either to GDP or to GDI appears optimal.