[jsun]

With absolute neutrality towards the topic of the article, here are some warning flags for the data presented.

"Employed Americans" isn't a typical survey category, and the author makes no attempt at explaining why "Employed Americans" was a more relevant demographic to survey than all americans, all adults, all adults between ages of x and y, etc. Nor does the author present equivalent statistics for the more commonly surveyed demographics.

"Nearly one in two employed Americans" really just means almost 50%, and could be misleading on quick glance.