PEN-L message, Query

I realize that my submissions generally don't measure up to the quality standards of the list and for that reason deserve to be ignored, but perhaps those in need of a little pro bono work might offer some enlightenment on the following perplexing matter.

The Economic Research Service of the USDA produces an abundance of data on the condition of US agricultural production.

I find particularly interesting that table on capital stock 1948-1999 which shows an approximate 50% increase in capital stock for the entire period, yet a real, and dramatic decline of some 33% between 1983 and 1999. Now this makes sense to me, given the "overweighted" portion of production value contributed by farms with sales greater than $1,000,000-- the ability of concentrated capital stock to be be smaller in volume, but denser in output and to "absorb" greater amounts of manufactured and farm based inputs.

However, when looking at the DofC BEA NEA tables for investment in and net stock valuation of non-residential, private fixed investments for farms, such valuations show no decline but an increase for the 1983-1999 period.

I'm having some difficulty reconciling the two, or even finding the paths of divergence.

Has somebody encountered the same issue and perhaps found an explanation?

Note to Sabri: The guy who knows what heteroskadastic (sp?) means doesn't understand obfuscation? That's precious.