One of the arguments for special government support of manufacturing is that manufacturing is a key to industrial R&D throughout the US economy (e.g.,
here). The data above comes from
BEA and
NSF and shows that as manufacturing declined as a portion of GDP from 2002 to 2007, industry R&D nonetheless increased (note: NSF data goes to 2007). Since 2007 manufacturing has continued to decline as a proportion of GDP (to 92 in 2010 in the graph above) while it appears that industrial R&D has tracked GDP (e.g., see
here in PDF).
With industrial R&D increasing faster than GDP over the past decade, even as manufacturing fell sharply as a proportion of GDP, I find little support for the argument that support of manufacturing is in some way critical for supporting industrial R&D.
