Is knowledge obsolete? The phone in your pocket can answer just about any factual question you care to pose. The cloud knows more than you do or ever will, and its store of knowledge grows with every passing second. So why should we bother to fill our heads with facts?

This is a question we’re all wrestling with, one way or another. There are many approaches to answering it. Traditionally it was held that there is a certain body of facts that every well-educated person should know. But the choice of these canonical facts is subjective. Canons are an increasingly tough sell in our diverse, digitally savvy society.

There are more objective ways of valuing knowledge. I examine several in my book Head in the Cloud. For instance, there is what I term a “knowledge premium.” People who have a higher level of general knowledge—demonstrated by answering quiz questions on survey—have higher average income. Often their self-reported health and happiness are better, too.

On the face of it, this isn’t surprising. Those with more education make more money (the unsubtle pitch for taking out those student loans). Affluence generally correlates with better health care and life satisfaction. The surprise isn’t the knowledge-income connection but rather that it remains substantial even when you factor out educational level and age.

I ran a national survey asking Americans to answer moderately easy sports questions. (What sport originated the term “full court press? Where does a shortstop play? How many players are on a soccer team?) Those who scored highest reported more than twice the household income of those who flunked ($85,245 vs. $33,969). These figures are adjusted to apply to a 35-year-old with four years of college. The difference is almost as big when you factor out gender.

Many other kinds of knowledge that correlate strongly with income, such as tests of history, correct pronunciation, current events, and finding countries on maps; and even a quiz I cribbed from the TV show Jeopardy!

Correlations (that don’t prove causation) are easy to come by our data-rich age. It’s harder to figure out what they’re telling us. The message seems to be that there is a real-world value to knowing things that goes beyond a college degree or the socioeconomic class associated with a given level of education. I suspect that one important thing knowledge gives us is context. This is helpful in evaluating new ideas and information.

Political scientists Kyle Dropp, Joshua D. Kertzer, and Thomas Zeitzoff offered devastating proof of that in 2014. As Russia’s invasion of Ukraine was in the news, they asked Americans to find Ukraine on a world map. Most couldn’t do it (no surprise there). The payoff was this: The further people’s guesses were from Ukraine’s actual location, the more likely it was that they favored a U.S. military intervention in Ukraine. Hawkishness correlated with geographic cluelessness.

War rooms have maps, for good reason. A country’s location on the map helps determine its strategic importance and the feasibility of deploying troops there. Oddly enough, some of the survey subjects picked points in the world’s oceans as the location of Ukraine. Either they didn’t know which parts of a world map were water—or they were counting on the submarine fleet.

I’ve found a similar effect with support for a U.S.-Mexico border wall. Those who scored lowest on a general knowledge quiz were more likely to say they favored a border wall. (One of the factual questions asked what country New Mexico is in. About 8 percent of Americans say Mexico.)

Border wall supporters have cell phones like everyone else. They can look up any fact they want—including whether the wall is supposed to go north or south of Albuquerque. But the survey measured what they had actually learned and retained, and the wall supporters knew fewer facts than others.

There has been much discussion of confirmation bias and its relation to political views. It is human nature to seek out information that confirms what we already believe or want to believe. Cable news channels like Fox News and MSNBC are held to be prime examples, basing their appeal on audience desires for confirmation. The Ukraine and border wall studies suggest a parallel (yet opposite) effect, one especially pertinent to those mobile devices we all carry around. We don’t fact-check ourselves when we are too ignorant to know what facts need checking.

Ukraine is half a world away, and the Mexican border is long. Those with a basic grasp of the world map know this, and that knowledge may immediately prompt questions about the practicality of a war or a wall. But for many Americans Ukraine is just a word they’ve heard on the news, not a landmass they can locate on a mental globe (much less a culture or a political state). Google Maps is not by itself a substitute for broad, contextual knowledge of geography.

Americans’ shaky grasp of geography probably has something to do with the traction the border wall idea has gained. It’s easy to understand the appeal of a putatively simple solution to the complex problem of illegal immigration.

Public policy needs to be grounded in the real world of facts. The cost of building, maintaining, and patrolling a border wall will be proportionate to its length. It’s not hard to find relevant information on the internet: the U.S-Mexican border’s length, or historical parallels like the Berlin Wall or the Great Wall of China (and how well they worked). But looking things up is easiest for those who have a sense of historical and cultural context. Without that, it’s difficult to craft an effective search, or even to know what you should searching for.

The internet is thus an instrument for leveraging knowledge. It increases information inequality: The smart get smarter, while the not-so-smart get left behind. That makes one of the best cases for the continuing relevance of knowledge. You need to know facts to look up facts.

William Poundstone is the author of 14 books, including Rock Breaks Scissors, Are You Smart Enough to Work at Google?, How Would You Move Mount Fuji?, and Fortune’s Formula. His most recent book is Head in the Cloud: While Knowing Things Still Matters When Facts Are So Easy to Look Up. He has written for The New York Times, Harper’s Magazine, The Economist, and Harvard Business Review, among other publications, and is a frequent guest on TV and radio. He lives in Los Angeles.