The “yes, college is worth it” side notes that reliable data sources, such as the Bureau of Labor Statistics, say education pays. And the College Board’s annual report, also titled Education Pays, consistently shows that those with college degrees earn more than those without degrees. Advocates on this side, such as Anthony Carnevale, get frustrated that anyone can see these numbers and not conclude that college is clearly worth it.

The other side looks at this same data and concludes “yes, but…” They have two main caveats they want acknowledged.

First, they note that many people who start college don’t finish college. This complicates the go-to-college argument because these students pay tuition, sometimes for years, but do not receive the average earnings bump from having a degree. There are many of these students – around 40% of four-year college entrants haven’t graduated after 6 years. If you just compare the earnings of college graduates to non-graduates, they are completely ignored. Thus, the high returns of college degrees are in part an illusion formed by selectively excluding students that had negative or low returns. This is like calculating the return on your 401k after throwing out all the investments that lost value. The resulting number may make you feel better, but it’s not an accurate depiction of reality.

Second, while on average college may still be a good investment, averages can hide a lot of variation, and there can still be some predictably bad investments. Consider the College Board’s Education Pays report, which shows that those with a professional degree earn on average $55,000 more per year than bachelor’s degree recipients. Does that mean that the answer to the question “is law school worth it” is a clear yes? Anyone who knows the carnage that is the legal labor market knows that those broader averages are not always representative, and give a very misleading sense of the likely payoff of going to law school.

And this brings us to why the ‘Is college worth it?’ debate is now obsolete at least as it relates to advice for potential students. Basing our advice on average earnings by level of degree is fine when that is the best data available. But those averages are no longer the best data available. The Department of Education recently began releasing earnings and debt data at the program level (e.g., the earnings for accounting majors who graduated from Ohio State). Students can find this data at the federal government website College Scorecard. My organization, The Texas Public Policy Foundation, has released a new webtool that presents that same data in a more user friendly and customizable format.

With this new data, it is clear that the “yes, but…” crowd was right that there is no one-size-fits-all answer to the “is college worth it” question. Some programs are worth it, and some aren’t. Averaging them together is a disservice when giving advice to potential students. With no consistent answer to the “is college worth it” question, a much better question is “which college degrees are worth it?”

As one method of answering that question, consider the debt-to-income test results from the TPPF webtool, which asks whether graduates from a program earn enough to repay the typical amount borrowed to attend that program. There are over 3,700 programs that failed this debt-to-income test, and almost 7,000 would be on probation. This means that there are hundreds of thousands of graduates every year attended programs that did not pass a debt-to-income test. My advice is that students should think twice about enrolling in these 10,000 programs. Following the “yes, college is worth it” crowd’s advice would send them off to college, where many of them would end up in these risky programs, like sheep being led to the slaughter. And that is the danger of asking the wrong question.