I've tallied number of positions, total ads and total area* in square centimeters for industry, governmental (foreign and domestic) and university positions. For the academic page, I've tallied the type of job. In the long run, I will do the same for the industrial positions by field (analytical, organic, etc.) and position (BS/MS, PhD., whatever).
1. Academic >>>> Industrial >> governmental positions.
There are a lot of tenure-track positions out there -- a lot. The average number of tenure-track professor positions published per week far outstrips total number of industrial positions offered. Governmental positions are few and far between.
2. September? Not so bad. December? Well... There are good hiring months and bad ones; September seems to be pretty good, while December is a terrible time to try to find a job. I suppose that's not such a big surprise. I know there's a cyclical nature to hiring in chemistry, I've just yet to figure out what that is exactly.
3. The market's not so great for working chemists. With the exception of the occasional mass ad, neither chemical companies nor Big Pharma have been seeking new employees. This is not good for either those of the layoff class of fiscal year 2008 or the new grads of academic year 2009. Then again, maybe there will be a turnaround and all of this will be moot. I don't think so, but one can hope.
*I've decided to use total area as the Chemjobber Index number from now on; if it proves not be useful, I'll change it. While it's not perfect, I think it's a good combination of the measure of intensity (willingness to spend money to try to find someone) and the number of positions available (while half-pagers can be used to try to hire one senior-level member, they're also used to hire many people with one ad.)