To call that chart incredible due to “cherry picked data” is sort of whistling past the grave yard.
Had another year in the late 70’s or early 80’s been used in place of 1981 the difference would likely have lowered the column only a few percent at best.
Had data for 2010 even been available in place of the column for 2009 the difference would probably not have been a percentage point.
You may well find the chart unimportant, but it is entirely credible.
A far better chart would have shown continuous data (say a line chart, with annual data). The events mentioned in the article could have been marked by vertical lines, at the corresponding years. That would have put the data into context, and would prove that it wasn't simply cherry-picked. Also, continuous data would have provided some information regarding trends.
The thesis of the article may, or may not be right. The chart does nothing to advance it, and actually makes it less credible (because of the obvious cherry-picking).
A far better chart would have shown continuous data (say a line chart, with annual data). The events mentioned in the article could have been marked by vertical lines, at the corresponding years. That would have put the data into context, and would prove that it wasn't simply cherry-picked. Also, continuous data would have provided some information regarding trends.
The thesis of the article may, or may not be right. The chart does nothing to advance it, and actually makes it less credible (because of the obvious cherry-picking).