One of my favorite modern historians, the late Rufus Fears at the University of Oklahoma, was fond of saying that “ideas make history.” University of Chicago professor and North Carolina native Richard Weaver expressed a similar view in his famous 1948 book Ideas Have Consequences.

Sounds obvious, I know, but both men were challenging the orthodoxy of scholars, usually but not always Marxists, who believed that inexorable social forces determined the outcome of history. Ironically, Marxism itself was one of the most consequential ideas in human history, producing suffering and death on a massive scale.

Another powerful idea, one with revelatory rather than destructive consequences, was that of diminishing returns. A related concept in economics, the theory of diminishing marginal utility, became influential at about the same time Marxism did, in the late 19th century. It was independently described and developed by three different economists in the 1870s: William Stanley Jevons in England, Carl Menger in Austria, and Leon Walras in Switzerland. Not coincidentally, they helped found three of the great modern traditions of economic thought: neoclassical economics, Austrian economics, and general-equilibrium theory, respectively.

What’s the idea? That you can’t predict the effect or value of something without first knowing how much of it is already present. In farming, for example, you get a lot of initial value out of fertilizing your fields. As you keep adding more fertilizer, however, each new increment has less of a productive effect than before. Eventually you reach a point at which additional fertilizer is actually harmful.

Marginal utility is a valuable tool for explaining things that might otherwise seem puzzling. Why can the same good be worth $5 at one time or place and only $1 at a different time and place? Because value is determined by circumstances. A lost traveler stumbling out of the desert will put a higher value on a drink of water than will an office worker who sits near a water fountain.

In public policy, the concept comes into play when discussing issues such as transportation and education. In the early 20th century, when the automobile was just beginning to become a popular consumer good, state and local governments used property and gas taxes to build their first real networks of graded, dedicated, and, later, paved roadways. The economic and social benefits of this investment were massive. Next, during the 1950s, a new wave of federal, state, and local spending gave us limited-access highways (interstates) and a more extensive set of paved secondary roads. Again the benefits were substantial, albeit not quite as massive as the first wave.

Today, the practical value of adding more lanes or highways to the existing road network is smaller. That’s not an argument against building them. It is an argument for being choosy about which ones to build, so that the marginal benefits remain likely to exceed the marginal costs (which are really the alternative goods we could purchase with the same dollars).

In education, there has also been a clearly diminishing return on tax dollars spent. During the early to mid-20th century, rising expenditures were associated with building schools where few existed, offering high school to many for the first time, equalizing opportunities across class and race, and establishing minimum standards for personnel and curriculum. Since the early 1970s, however, real, per-pupil spending on public education risen substantially while the average performance of 17-year-olds on national math and reading exams has barely budged.

The storyline is a bit different in North Carolina, where additional programs and educational expenditures during the 1970s and 1980s may partially explain the significant gains that North Carolina students made on independent reading and math exams during the 1990s. But the next wave of state reforms and spending increases, during the 1990s and part of the 2000s, did not produce a commensurate rise in achievement among the affected students.

In neither transportation nor education has North Carolina reached the end of potential improvement, of course. The point is that additional progress will have to come largely from productivity gains, not from simply spending tomorrow’s money on yesterday’s terms.

John Hood is chairman of the John Locke Foundation.