Last month I attended a Federal Reserve Bank and Upjohn Institute conference on expanding opportunity within workforce development programs. It isn’t possible to review the breadth of the research and programs I heard about, but I think many readers will appreciate one issue that raised its ugly head throughout the conference—the role of state departments of labor in distorting labor markets and poorly projecting skill needs. These issues offer a great deal to write about, so let me just focus on two matters—education and ‘labor shortages.’

Since 1990, the United States has not created a single net new job for workers who have not been to college. Worse still, wages for non-college attendees are now lower than they were in 2000. So, even at full employment, labor market outcomes for workers who have not been to college are, on average, terribly poor.

The situation in Indiana is about the same. We haven’t created a single new job for high school graduates in 20 years, which is as far back as we have reliable data. Though high school drop-outs have seen some job growth in recent years, it is only enough to account for 1.8 percent of jobs. Here, too, wages are collapsing, and now sit at less than 75 percent of the 1979 level after adjusting for inflation.

In the face of these US Department of Labor data, Indiana’s Division of Workforce Development tells us that between 2014 and 2024 Indiana will have to fill 1 million jobs, most of which will require only a high school diploma. With three years of a booming recovery behind us, it is safe to conclude that these predictions are profoundly mistaken in both magnitude and composition of new jobs.

To be fair, all forecasts are wrong, mine included. The real question is whether they are correct enough to be useful. Here also, the answer is a resounding no. I’ll use one of our more popular training programs to illustrate why.

Trucking companies nationwide complain of a truck driver shortage. Looking at employment data, you’d be inclined to agree with them. Total employed truck drivers have just now returned to their pre-recession level, at a time when they should probably be 15 percent higher based on the flow of goods. However, employment alone does not tell us if there is really an excess demand for truck drivers. To determine if demand is growing faster than supply, we’d need to look at wage data.

As it turns out, inflation-adjusted wages for truck drivers throughout 2017 grew slowly and were lower than they were in 2001. So, despite all the teeth gnashing and crying, there is cannot be a real excess demand for truck drivers. If there were, wages for truck drivers would be much higher and rising.

Now, these aren’t my facts; they belong to the Department of Labor. What they state clearly is that the labor shortage in truck drivers is just like the Lamborghini shortage I face at home. I want a Lamborghini, and my teenage sons want me to have a Lamborghini, but I don’t want to pay the price for a Lamborghini. Instead, I drive a 10-year-old pickup truck and face a terrible shortage of Lamborghinis in my garage. This, I am quick to note, is not a public policy problem.

So, one might ask, why does Indiana pay so much to train new truck drivers, and list truck drivers as a “Hot 50 Job?” In terms of job growth, truck drivers ranked 187th and 191st out of roughly 500 Hoosier occupations since 2014. In terms of wage growth, they ranked 130th and 395th statewide. Again, these are US Department of Labor data, drawn right off the website this week.

These shocking facts ought to motivate a statewide policy debate. After all, we are allocating a significant amount of state resources into training truck drivers and telling young Hoosiers that these are ‘Hot Jobs.’ But, there’s an even more unpleasant problem. I believe there is evidence that our training programs are actually suppressing wage growth. Since 2014, fewer than a quarter of all “Hot 50 Jobs” requiring only a certification in Indiana have had inflation-adjusted wage increases. I don’t know if this is simply the result of wretchedly poor occupational projections or if we’ve pushed so many young adults into these careers that it is quelling wage growth. Either way, our Workforce Development efforts are cruelly failing a number of Hoosiers.

There is more to worry about, like a complete lack of focus on automation-related job risk. But, if we cannot even get our ‘Hot 50 Jobs’ anywhere near right, preparing for the labor market effects of artificial intelligence is a pipe dream. The long-term fix to this problem isn’t going to come out of a committee meeting or employer focus group, but out of solid, uncompromisingly honest data analysis. Our workforce development system has fallen well short of that goal.