I would like to draw your attention to several problems with the article "Employment Growth Is Leveling Off" (ENR Construction Facts, November 2003 p. 62). The introductory paragraph discusses construction trends for 1993-2000 and 2001-2003. In contrast, the table "How Construction Union Membership Is Changing" is for an entirely different time period, 1995-2000.
Instead of allowing readers to draw conclusions about the strength of the construction industry versus the strength of construction unions, the mismatched time periods only serve to mislead the reader and allow the data to be misused.