Over the past 30 years, companies in the United States have added 40 million jobs. This is a remarkable achievement that enriches Americans and dramatically improves our standard of living. As we think about this growth, it is helpful to consider what type of jobs have been created, what types of workers are hired and what these facts tell us about the types of economic policies we pursue at the state and local level.

I’ll begin by dividing all job types into two categories. The first are ‘global’ jobs. These are jobs that produce a good or service that is sold mostly outside the region in which it is produced. Think of manufactured goods such as automobiles or RVs, agricultural commodities, petroleum products, software or medicines. The ‘global’ definition also includes services such as website design and development, back office financial services and, increasingly, medical diagnostics.

The other type of job is a ‘local’ job. These jobs produce goods and services that are consumed locally. So, preparing a meal in a restaurant, mixing cement, teaching school or litigating a court case are done primarily for local consumers. So are physician and nursing services, lawn care and home construction, retail, childcare and grocery sales.

Of course, not all jobs fit neatly into these categories; maybe 2.0 percent of jobs could fit either definition, but that doesn’t really change the overall point I’m trying to make. Also, by local I mean a product or service is sold in the same city or in the same and neighboring counties. This means I’m counting a GM assembly plant worker as having a ‘global’ job even if she buys a GM car.

Over the past 30 years, the U.S. has lost between 2 million and 7 million ‘global’ jobs. That means of the 40 million net new jobs in the U.S. since 1992, a total of between 42 million to 47 million are ‘local’ jobs. No U.S. state, and very few cities, have more ‘global’ jobs than they did in the spring of 1992.

Chasing these ‘global’ jobs has been the primary economic development policy for many state and local governments for the last 50 years. Every tax incentive, every tax abatement, every business-friendly tax cut or regulatory change is couched in terms of attracting these businesses. So, on the face of it, these are hundreds of billions of tax dollars spent over the past 30 years chasing a declining number of jobs.

This pursuit is puzzling, but not as bewildering as the fact that most of the policies that focus on attracting business shift resources, mostly tax dollars, away from the very things that attract the ‘local’ jobs. These are good schools, clean and safe neighborhoods, parks, and the like.

To be fair, most successful state and local governments have shifted focus away from attracting ‘global’ jobs to nurturing their ‘local’ jobs. Sadly, many economically unsuccessful places are still chasing the rapidly disappearing ‘global’ jobs. In fact, one way to measure the economic vitality of region is simply to measure how much money is spent attracting jobs. The more resources a community spends in trying to lure ‘global’ jobs, the more unsuccessful it is, and the more desperate it becomes.

All of this primes the question, who is getting these new ‘local’ jobs and how well do they pay? When I was a young lad, many Americans disparaged ‘local’ jobs. They were primarily service sector jobs, as they are today. Back then, one in three Americans worked in factories. Today, it is one in 13, and the share continues to shrink.

The great American employment growth of the past 30 years has been in these ‘local’ jobs. As it turns out, these new jobs have gone overwhelmingly to college graduates. Nationwide, over the past three decades more than 8 in 10 net new jobs went to folks with a four-year college degree or more. That is over 32 million jobs.

The remaining 8 million jobs went to folks who earned an associate’s degree or had some college training. Sadly, the data don’t distinguish between the two groups. Folks without any college training or have less than a high school diploma comprise fewer jobs than they did in 1992. American companies hire fewer workers without college than they did 30 years ago. This doesn’t mean there aren’t good jobs for people who don’t go to college. Plumbers, electricians, carpenters and welders retire every year, which creates continual turnover in these sectors. There just won’t be growth in these occupations.

However, the declining demand for workers without a college degree means that there won’t be net growth in these sectors. This is not new. Demand for college graduates has been rising for 75 years, while demand for high school graduates has been in decline over the same time. It is safe to say that nearly all the net job growth over the next several decades will be for college graduates.

Of course, some have argued that many of these college graduates are underemployed or holding jobs that don’t require a college degree. That’s an issue to take seriously. Fortunately, it is a hypothesis that is easy to test, simply by looking at wages. Businesses won’t pay more for education they don’t need, so if there is widespread underemployment among college graduates, it should be visible in wage data. Increasing or widespread underemployment will mean lower wages for college graduates relative to everyone else.

It turns out that not only is nearly all job growth concentrated among college graduates, that is where most of the wage growth occurs as well. Over the past 30 years, the wage premium for college graduates relative to high school graduates has risen by more than 5.0 percent. In the 21st century alone, 82.4 percent of inflation-adjusted wage increases that accrued to people with a high school diploma or higher went to college graduates.

What does all this mean about economic development policy for states and cities? I think the first lesson is that places that focus chiefly on attracting business should better prepare themselves for continued disappointment. Second, the future of wage and employment growth belongs to regions with high levels of college graduates. States and cities that wish to grow their economy will have to focus more on educating and keeping these workers.