The economic view of growth is pretty simple. People and businesses are attracted to places based on three factors—local productivity, quality of life and the responsiveness of housing supply to population growth. Every regional economics class in an American university will cover these three topics. Students will learn the reasoning behind this model—the math theory behind the ideas. They’ll then read the empirical research—the statistics that test the theory. This is my summary.

Productivity differences between places in the U.S. are explained almost exclusively by differences in educational attainment and population density. Well-educated, high-density cities and towns are very productive. That means their GDP per worker, or GDP per resident, is very high. Decades ago, the amount of business capital a region possessed drove much of productivity differences. So, in the 1950s, Indiana’s factory-rich landscape made us a relatively productive place.

Today, the link between business capital and regional productivity has largely disappeared. We now rank between 32nd and 40th in productivity, depending on the measure. To be sure, capital still matters in less-developed nations, but in the developed world, brainpower, not machine power, has been the dominant source of productivity differences for half a century. So, policies that nurture human capital produce more productive places.

This change continues to confound many policymakers, who remain wedded to the belief that prosperity is just one new factory away. It reminds me of Keynes’ famous quote on that phenomenon, “Practical men who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist.”

A half century ago, productive places typically had low quality of life. That is changing, and today there is a convergence between highly productive places and places with high quality of life. So, most households and business no longer need to choose between a place that is good for business and good for families. Successful places increasingly do both.

Measuring quality of life is a bit more difficult than measuring productivity, but it is as conceptually easy. The most common approach involves two steps. First, create a statistically identical house in each city by controlling for all the specific home features like number of rooms and year of construction. Then, measure how much more or less people are willing to pay for that home than its construction characteristics suggest. That is a measure of how much people value these homes. If they pay more, it is a nice place. If they offer less, it is less appealing.

The second step is to do the reverse for labor markets. Create a statistically identical worker in each city by controlling for education, age, occupation and other factors. Workers who demand a wage premium to live in that city may find it unattractive. If they are willing to accept a lower salary, they may prefer the location. Combining these two measures yields a county or citywide measure of quality of life.

This measure of quality of life indicates where people prefer to live, but it does not tell us what they prefer. The best way to ask that question is simply to find which amenities best predict the quality of life measure from housing and labor markets. Several co-authors and I have done this work across the nation and in the Midwest. The results are partially predictable, but they also hold some huge surprises.

Natural amenities like mountains, lakes and rivers are appealing, but have only a modest effect on household and business location decisions. Weather matters, but again, not very much. The migration to warm places that accompanied the invention of air conditioning appears to have run its course. Families do prefer private amenities, such as good groceries, fitness centers, recreational facilities and the like. Here, variety seems to matter substantially.

The biggest predictor of quality of life differences across the country are public services. The share of the economy allocated to education is the number one attractor for families. Number two nationwide is crime rate. Good, well-funded schools and low crime rates explain more variation in quality of life than does all natural amenities combined. Yes, parks, broadband, nice restaurants and mountain views matter, they just matter far less than safety and good schools.

The final puzzle to all this is the responsiveness of housing supply to population change. Places that grow their housing supply with population change tend to do much better. At the same time, places that shrink their housing supply to match population decline tend to do better. This symmetry is as frequently ignored as it is important.

This simple framework explains most differences in economic growth between places in the United States, and has been doing so for half a century. The policy implications of this are gobsmackingly clear. To grow, cities should push educational attainment as hard as possible, ensure families have safe places to live, and create an environment that allows new businesses to meet customer demand for goods and services. It stands to reason then that this view would dominate state and local economic development policy.

For Midwestern states, the urgency of this is even greater than for elsewhere. Our economies were built on the notion of capital-intensive industries that attracted workers from around the country. That is no longer the case, yet the muscle memory of capital incentives dominates the way we think about economic growth. It has left us saddled with a history of ineffective policy approaches.

Even when we give lip service to quality of life, there is a surprising resistance to investing in those things that most influence household and business location choices. I suspect most of this opposition is driven by the belief that low-tax and low-spending efforts will attract more business capital and that this in turn will boost productivity. This view is sadly mistaken.

Indiana’s workers are only 74 percent as productive as those in high-tax California, and unsurprisingly, earn only 74 cents for every dollar they do. Ohio, Wisconsin and Michigan find themselves in nearly the same circumstance. I know everyone likes to ridicule Illinois, but their workforce is 15 percent more productive than Hoosiers, and their wages 15% higher.

It is past time to suppose that low taxes and business investment drives economic growth in the developed world. Instead, it is educational attainment. Quite simply, the share of adults with a bachelor’s degree or higher explains a full two-thirds of productivity differences between states.

Capitalism has drawn more people from poverty than all other economic systems combined. It is unabashedly the economic system of prosperity. Two hundred years ago, productive land was the most important form of capital. A century ago, business capital was the most important. Today, human capital is what matters. Economic development policies that focus on human capital will yield productive places, with good quality of life and responsive housing markets. Anything other than this will continue up to disappoint.