Weapons of Math Destruction: Chapter 3: Arms Race Summary & Analysis

To explain one of the core components of a WMD—scale—O’Neil invites her readers to imagine that the trendy “caveman diet” became the national standard, and all 330 million Americans were forced to follow its dictates. The restrictive diet, which favors meats, fish, fruits, vegetables, nuts, seeds, and cheeses, would have a huge effect on the economy—if it became a national standard, it would create a distorted economic climate. This is precisely what has happened to higher education.

In 1983, the U.S. News & World Report, a struggling magazine, decided to evaluate and rank 1,800 colleges and universities across the U.S. to bring in readership. They based their rankings off opinion surveys sent to university presents—but after the first rankings were released, complaints started pouring in, and the editors at U.S. News tried to figure out how they could statistically measure the vague concept of “educational excellence.” So, they decided to look at SAT scores, acceptance rates, student-teacher ratios, alumni donations, and more, building an algorithm to create the first data-driven ranking in 1988.

The rankings, though, quickly became mired in a feedback loop—low-ranked schools received fewer applications from top-tier students. Thus, their percentage of high-achieving students dropped, as did their revenue from admissions and alumni donations. And the U.S. News list kept growing in reputation and scope—soon, it was a “bona fide WMD.”

Many schools began looking for ways—both legal and extralegal—to bump themselves up in the rankings. But the U.S. News model was constructed from proxies—the most relevant data about students’ day-to-day experiences at their colleges wasn’t accessible. So, it was easy for colleges to improve in the arbitrary areas the ranking measured. But as universities hustled to improve faster than their competitors, a kind of “arms race” began. Texas Christian University poured money into its student center and its football team to attract more students. Their strategy worked—by 2013, it was the second-most selective university in Texas. It climbed 37 places in the ranking in just seven years.

The rankings were creating a kind of rat race in U.S. academia. As more and more schools adopted algorithmic approaches to things like admissions, safety schools started rejecting the top-tier candidates whom they knew would likely reject offers for better universities. So, safety schools were no longer “safe.” O’Neil argues that it’s not just students who are suffering in this new climate—it’s the schools, too, who are losing out.

By leaving tuition, fees, and student financing out of their initial model, O’Neil argues, U.S. News ultimately did its readers an enormous disservice. By ranking expensive, prestigious universities highest, they implied that money was the cost of excellence. And indeed, between 1985 and 2013, the cost of higher education rose by over 500 percent—four times the rate of inflation. And while the student loan crisis isn’t the fault of U.S. News alone, the idea that a degree from a highly ranked school is a path to power and success is partially their doing.

Colleges manage student populations “like an investment portfolio,” according to O’Neil, by assessing which students are assets (those who pay full tuition or have families who donate money) and which are liabilities (student athletes who receive lots of scholarship money to attend). Education consulting firms have sprung up to do the analytical work of forecasting and ranking enrollment prospects by a number of categories and characteristics—often using the U.S. News metrics as a model. All over the U.S., rankings—and efforts to game them—continue to grow and spread.

To try to level the playing field, Saudi universities paid successful academics exorbitant fees to come on as adjunct faculty. And students in rural areas of China protested in defense of their right to cheat on standardized tests (to even things out between them and their urban counterparts). It was becoming clear that the rankings had created a system that many people felt there was no way to win fairly. Expensive application bootcamps and admissions coaching sessions now dominate education by turning students' test scores and GPAs into statistics. And colleges and universities create admissions models that are hidden, large-scale, and trapped in feedback loops—they are WMDs. The contemporary education system favors the privileged—those who have the means to play by the algorithms’ rules. 

During President Barack Obama’s second term, he suggested a new college rankings model that was tied to affordability, diversity, and postgrad job placements—but any ranking system, no matter its priorities, can still be gamed. For instance, a law school graduate with hundreds of thousands of dollars in loans working as a barista is considered employed. Another example is that schools can lower costs by replacing retired tenured professors with overworked adjuncts who cost less to hire.

The government failed to rewire college rankings. Instead, the Education Department released lots of data online, hoping that students would use the data to determine what matters to them individually about their collegiate experiences. The website allows people to create models for themselves that are transparent, user-controlled, and personal. O’Neil suggests that the Education Department’s new site is “the opposite of a WMD.”