The Wall Street Journal reported yesterday that while manufacturing has experienced some modest job growth lately, manufacturers are struggling to fill their open jobs. The story notes that the number of open manufacturing jobs has steadily risen since 2009, and that openings in manufacturing are at their highest level in 15 years.

I’m always a little suspicious, though, when anyone starts the ticker in 2009, which was when we were in the middle of one the worst recessions in history. Pretty much everything looks better if you start in 2009. It’s worth pulling the data from further back in time to get a better a better perspective on this recent growth:

Total U.S. Job Openings, Manufacturing, Dec. 2000 – Jun. 2016 (In thousands, seasonally adjusted). Source: Bureau of Labor Statistics

Here you can see that, yes, manufacturing job openings have rebounded since 2009, although they really had nowhere to go but up. In fact, they’ve climbed back to just about where they were in 2006. (We hit a high of 396,000 openings in April of that year, and there were 397,00 openings in April of this year.) So actually they are possibly at their highest levels in ten years, not fifteen.

But equally important is just how many more openings there were 15 years ago, if that’s what you count as your high-water mark (the BLS data only goes back as far as December of 2000). In January of 2001, there were an estimated 496,000 open jobs in manufacturing—100,000 more than there are today.

The second thing that I think is worth pointing out is that while job openings in manufacturing have been on the rise recently, employment in manufacturing has been on the decline for decades, and I don’t think anyone sees it coming back to where it was a few decades ago. Here’s another chart from the BLS showing the number of people employed in the manufacturing sectors since 1979:

All Employees, Manufacturing, Dec. 2000 – Jun. 2016. (In thousands, seasonally adjusted). Source: Bureau of Labor Statistics

In this instance, the BLS data goes back to the 1930s, but I started at 1979 because this is the high-water mark for manufacturing employment in this country. (And remember, there were fewer people in the labor force back then, and, well, fewer people, period, so the percentage of people working in the sector was much higher than today.) But you can see that manufacturing jobs started rapidly disappearing in the 2000s, and really cratered during the Great Recession. So the sector has come back a little bit, but it’s nowhere near where it once was.

Executives told the Journal that the reason that jobs are going unfilled is because they can’t find workers with the skills to run the advanced machinery these companies have invested in. (The machinery that in many cases allowed them to lay off all of their lower-skilled workers to begin with.) But let’s say that  we did give all those lower skilled factory workers who got laid off over the last 15 years the skills employers say they need—I think it’s unlikely, looking at the chart above, that the sector is would produce enough new jobs in the foreseeable future to hire them all back again.

(Let’s table, for the purposes of this discussion, the argument as to whether, in certain cases, employers have a moral responsibility to do more than just hang up a “Help Wanted” sign and hope for the best in those communities that were hit hard when all that advanced machinery enabled them to cut a bunch of jobs. Some might argue that these employers ought to stop complaining and invest in what is needed to skill up those folks for these new openings. I’ll let you make up your own minds about that.)

It’s also worth noting—again—that there are those who study this stuff who argue that the skills gap in manufacturing  is overblown. I’m reminded also, of this paper, published a few years ago by the Federal Reserve Bank of Boston, which found that as a result of the Great Recession and the high unemployment rate that followed, employers simply got used to being able to hire workers with relatively high skills at a fairly low wage, a trend the study authors called “opportunistic upskilling.” Once the labor market tightened up again, hiring and retaining a workforce of higher-skilled people across the board was clearly going to be more challenging. They expected employers to adjust, by either lowering their skill requirements for some of these jobs or by raising wages. Those unwilling to do so would find that the average time to fill a position would grow longer. I have no idea if that is what’s going on at some of the companies highlighted in this article—but it could be!

The point of all of this is—it’s complicated. I don’t have any doubt that many manufacturing jobs require more advanced skills than they did a decade or two ago, but a reluctance by employers to raise wages, or to look at lowering the skill requirement for some jobs (and/or investing in more training on the job) may also be factors.

Why do I care about this? Well, because from a workforce training/adult education perspective, I want know with as much precision as I can what the actual employment trends are in order to figure out what the actual need for training really is, and whether there are other things we need to look at in order to get people into (or back into) a job.