I take issue with this math. The first set deals with percentage but ignores overall numerical values. It's the classic 1 -> 2 = +100% while 1,000 -> 1,500 = +50% error. Most new jobs won't necessarily come from the most percentage growth. The numerical value analysis is then isolated to this percentage growth (thus probably ignoring the 1,000 -> 1,500 industries).<p>For evidence of this, there's an est. +35mil influx of population from 2010 - 2020. If we assume that ~25mil of the influx will be working in 2020, then we have a much more modest ~8.4% growth (from the 2.1mil cited), not the infographic implied +75%.<p>But that isn't the whole story. There's buggy-whip industry syndrome (ie, negative growth industries): In the next 8 years there will jobs lost to automation; eg., bank tellers yielding to online banking, toll road operators yielding to fast-pay devices, you may even see driver-less taxis by 2020. Many of those people will be forced to change industry.<p>Since those negative number industries have to be accounted for, just looking at the influx of population is erroneous when considering potential positive influx. Also, the median age is set to increase, so that has to be accounted for as it implies that a larger percentage of people will be of working age.<p>Napkin math is going to get you in trouble when dealing with this kind of stuff since those things are much harder to estimate, but the 2.1mil cited accounting for 5% of jobs may be a much more reasonable assessment; or it may be 15%, but it's not 75%.<p>Don't get me wrong, I like the color scheme: nice fonts, good layout; I just find the lensing of the data to be manipulative.<p>When you have bad stats you have bad conclusions, a misinformed public, and bad public policy. That can't be good.<p>Besides, claiming a more honest figure of 1/10 of new jobs will be elderly care related doesn't detract from the message of industry's significant growth - it's still a really substantial slice of the pie.