From Media Matters:
It is difficult to understand how so many errors could be crammed into
one simple chart, merely by accident. It is even harder to
understand how the artist could end up with a straight line, after
incorporating numerous errors, unless it was an done with the intention
of being misleading.
First, (problem #1) the Y-axis does not start at zero. That is a
common error. Sometimes it is done innocently, and a notice is
posted. For example, “Y-axis does not start at zero, to better
show the change,” or something like that. OK. That’s minor.
The Fed does the same thing.
Notice, though, that the title says “Job loss by quarter.” But
(#2) the X-axis does not show quarters. The X-axis also is not
linear (#3). December 2007 to September 2008 is 9 months (3
quarters), while March 2009 to June 2010 is 15 months (5
The biggest problem, though, (#4) is that the numbers on the line do
not indicate job losses per quarter, or per any
As Media Matters points out, we clearly did not
loose 15 million jobs in any one quarter. Rather, those number
appear to represent the total number of unemployed persons in the USA
that are imputed from the BLS survey data. These are total
numbers, not losses per quarter.
This motivation for this whole thing is difficult to understand.
Yes, the job
situation is dreadful. Nobody needs to make stuff up, in order to
make that point. The simple truth is sufficient.
Yes, the rate of job loss has declined. There may even be some
net job creation going on. But even the most optimistic numbers
still show that job creation is lagging population growth. We
know this. We know it is really, really bad. We do not need
deliberately misleading graphs to establish this.
At first glance, it seems to be a simple case of incompetence. If
so, then the number of unemployed should soon be 15 million plus one —
once the person who made the graph is fired. But even an
incompetent person could have simply pulled href="http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=UNEMPLOY&s%5B1%5D%5Brange%5D=5yrs">the
chart from the Fed website. Not only would it be easier, it
would be more accurate.
But no. The most likely explanation is that they wanted a simple,
visually punchy graph that would convey a message at a glance, and would
not invite closer inspection. The Fed graph is a bit more
complicated. Someone might be tempted to inspect it and think
about it. Maybe they would even go the the site mentioned at the
bottom, and start poking around, start thinking about how they would be
better off getting their news somewhere other than Fox. After
all, the Fed graph shows that job losses stabilized shortly after we
got a new president.
Perhaps the intention is to somehow imply that the job situation has
worsened, and that the worsening could be addressed by electing
different politicians. If so, it is an incredibly naive