ADHD diagnosis rates have risen dramatically since 1983. Clinically measured levels of attention in random samples of normal kids, however, are unchanged according to a 2012 paper.
Indeed, the total lack of change in the result (from 2002-2006 test administrations) is itself surprising given the sample size, which all other things being equal would be expected to deviate just a little from the prior result due to the statistical noise associated with random sampling errors.
Maybe this means that kids who had ADHD before are now being properly diagnosed instead of denigrated. The clinical test shows about two standard deviation differences between typical kids without a diagnosis and those diagnosed with either ADHD or an autism spectrum disorder, which suggests that the diagnosis process today is far from random.
An increased diagnosis rate may also mean that more typical kids are being diagnosed when they should be and are receiving drugs to treat the condition that are counterproductive. The two possibilities are not only not inconsistent, but are both likely to be true.
But, the result undermines theories that claim that a greater diagnosis rate for ADHD is in any way a product of greater incidence rates of the condition.