Perpetuation of Privilege — Debunking the U.S. News Best Colleges Lists
Illustration of students walking on a college campus

Perpetuation of Privilege

Debunking the U.S. News Best Colleges Lists

It's the news drop heard around the world. Every fall, U.S. News releases its rankings of top colleges in the United States, and millions flock to the site (more than 10 million in the first 72 hours after publishing in 2007 (Chartier and Peachey, 2014)).

High school students and their parents rely heavily on the rankings to help them make the expensive and life-impacting decision of where to apply, and ultimately attend, college. One college senior recalled his time researching colleges, saying:

Parents and students look at the [U.S. News] rankings and assume they [U.S. News] did their research and know what they're talking about, so we don't have to.

Universities collectively spend billions of dollars and countless hours toward improving their rankings resources that could be directed toward student-benefiting programs.

The cost of college education has increased 140% since the U.S. News Best Colleges rankings were first published in 1983, with student loan debt reaching an all-time high of $1.75 trillion in 2021, outpacing the growth of the U.S. economy six times over (Hanson, 2021).

A line graph showing that the average cost of a four-year degree has increased 140% since the U.S. News Rankings were first published in 1983.

Following the COVID-19 global pandemic, during which many college students were forced into virtual learning environments, the education industry is facing serious disruption. Students and parents alike are questioning the value of their investment. Universities must justify the high price tags and consider how costly practices, like vying for placement on the U.S. News rankings, not only contribute to higher student costs and indebtedness, but also prop up questionable reporting practices.

In theory the U.S. News Best Colleges rankings measure the quality of academic instruction and experience.

But is U.S. News measuring what students care about? And how accurate and reliable is the U.S. News analysis?

Digging into the history and the calculations underlying the rankings, however, calls into question the value of the rankings.

The U.S. News Best College rankings were first published in 1983, when they were created to compete with Time and Newsweek, competitors that have all but disappeared today. The first rankings included 25 national universities and 25 national colleges, and the rankings were developed based solely on a "marketing survey that asked college presidents to rate the nation's top schools." (Paterno, 2021). President emeritus of Northeastern University, Richard Freeland, shared that, "Most of the people who were filling out those surveys had no idea what they were talking about." (Paterno, 2021)

In 1988, U.S. News introduced quantitative measures into the rankings, and those have changed almost annually since (Bastedo & Bowman, 2010). From an initial weighting of 100%, the Peer Assessment (Reputation) Score is now weighted at only 20% — though this remains the most heavily weighted individual factor, and one that receives a lot of scrutiny. One reason for this is that over time, the "U.S. News rankings [have] become college reputation" (Bastedo & Bowman, 2010) — some have gone so far as to suggest that the people completing the survey use the prior year rankings to complete the survey.

The rankings, aimed at consumers, were first placed online in 1993. Over the years the rankings have expanded to include more colleges and universities (over 200 in some categories) in each category, and more categories (national universities, national liberal arts colleges, regional universities, regional colleges, undergraduate business programs, and so on). U.S. News has even expanded to ranking elementary and middle schools this year.

This expansion speaks to the business of the rankings: more lists means more press, more hits online, and more ways to make money through the paid College Compass subscription. College is a "high-stakes economic and social transaction" (Paterno, 2021) with life-long implications; how reliable are the rankings so many use to make this choice?


The rankings don't align well with student priorities.

U.S. News includes 17 "factors" in its Best College rankings calculations, many of which are financially oriented.

There is limited overlap between what U.S. News factors into the rankings and what students applying to college care about:

  • Reputation is measured by U.S. News through a "Peer Assessment Survey" of the president, provost and dean of admissions of the other schools in the list (i.e., the 200+ other schools considered Liberal Arts schools) and high school guidance counselors. Arguably, the reputation among recruiters or hiring managers would be more meaningful.
  • Academic quality is indirectly measured through the four metrics in the table below.
  • Affordability is indirectly measured through graduate indebtedness. Average net price would likely be a better metric.
A table comparing what college applicants care about and what U.S. News incorporates into its ranking system.

U.S. News only addresses students' concerns about reputation / academic quality and affordability, and most metrics related to this are measured indirectly. U.S. News fails to address students' other concerns, while valuing metrics that students don't seem to care about.


The rankings are correlated with wealthier schools and students.

Many of the metrics U.S. News incorporates into the rankings are directly or indirectly related to how much money the school and its students have:

  • Graduation and retention rates (22% of the U.S. News Score) can favor wealthier students who have more resources at their disposal for tutoring and the like, and whose familial financial situation is less likely to negatively impact their ability to stay in school.
  • Faculty resources (20%) is tied to a school's financial resources and financial resources per student (10%) measures a school's spending.
  • Graduate indebtedness (5%) is likely to be lower if a student comes from a wealthier family, while alumni giving rate (3%) is likely to be higher for wealthier alumni.

The way the U.S. News ranking system is constructed rewards (intentionally or otherwise) schools with more money and who admit wealthier students.

A scatterplot graph showing that higher endowment is generally associated with a better U.S. News ranking.

A college's endowment is the money the school has raised from donors that it invests to support its educational and research mission in perpetuity (American Council on Education, 2021).

Though U.S. News does not directly incorporate endowment into its calculations, there is a correlation between endowment and the U.S. News rankings. More money in the bank generally is associated with a higher ranking, despite the fact that the amount of money in the bank shouldn't directly impact educational outcomes.

Endowment may also be correlated with how much money a college is willing to spend to try to boost its ranking, with some schools spending hundreds of millions each to improve their rankings.

A scatterplot graph comparing the U.S. News factors to endowment (wealthier schools) and Non-Pell Students (wealthier students).

U.S. News college rankings correlate strongly with student and college wealth.

Generally, the higher the wealth, the higher the ranking.

The animated scatterplot to the left shows the correlation between each U.S. News factor and endowment (horizontal placement) and Non-Pell Grant Students (vertical placement). Students receiving Pell Grants come from families with lower income.

In his Revisionist History podcast episode Project Dillard (about an Historically Black College or University (HBCU), Dillard University), Malcolm Gladwell describes the STEM (science, technology, engineering and mathematics) "leakiness" problem that exists at most schools:

An illustration of Malcolm Gladwell, with a quote about STEM pipeline leakiness at most schools compared to HBCUs where the pipeline isn't leaky.
0:00

Malcolm Gladwell describing the STEM leakiness phenomenon.

Educational researcher, Mitchell Chang, explains that students who are in these challenging STEM programs who don't make it to the top of their college classes lose their confidence and abandon their pursuit of a STEM degree. However, HBCUs, compared to more selective institutions, are more focused on helping their students succeed in STEM programs.

An illustration of Mitchell Chang, with a quote about the difference between what HBCUs do versus predominantly white institutions in developing students to achieve their potential.
0:00

Mitchell Chang about how HBCUs combat STEM (science, technology, engineering and math) leakiness (when students enter into a college as STEM majors but ultimately "drop out" of those majors because they are "too hard.")

A bar chart showing black physics graduates from Dillard University and Harvard University. Dillard has graduated 35 black physics graduates in the last 10 years to Harvard's 5.

Despite providing excellent education at a great value, HBCUs are for the most part excluded from the U.S. News Best Colleges list. Only four HBCUs made it into U.S. News' Best Liberal Arts Colleges ranking since 1983.

An illustration of Walter Kimbrough, with a quote calling the U.S. News rankings a perpetuation of privilege.
0:00

Walter Kimbrough about the biases inherent in the U.S. News Best College rankings and how they impact colleges that serve underprivileged communities.


The rankings and data are not transparent.

U.S. News portrays itself as being transparent, releasing an article with the rankings each year describing "A More Detailed Look at the Ranking Factors." Despite this detail, there are gaps in the data.

The factor weightings are not fully transparent.

U.S. News presents a table of the factor weightings, but the descriptions show that the full formulas are still hidden.

A table of the U.S. News Best College ranking factors and their percentages, showing that four class size factors have undisclosed individual percentages that add up to 8%.
The class size factors add up to 8%, but the specific weightings are not disclosed.

U.S. News does not disclose all of the data used to calculate the rankings.

A detailed bar chart showing 301 missing data points in U.S. News' College Compass database, broken down by highest weighted factor.

U.S. News sells annual subscriptions to its College Compass service, which supposedly contains the data used to produce the U.S. News Best Colleges rankings. However, there are 301 (8% of the total) missing data points throughout.

Some schools (30 of the ranked Liberal Arts Colleges, to be exact) did not submit data to U.S. News for the 2022 rankings, which could theoretically explain some of the gaps.

However, some of the data points (like graduation rates for students who received Pell Grants, for which 14.8% of the data was missing) are available from the government IPEDS database.

IPEDS is an annual government-sponsored survey that colleges are required to complete if they are eligible for federal funds.

An illustration of the formulas that U.S. News uses to incorporate standardized test scores into their Best Colleges Ranking, next to the data U.S. News provides, which is not what it uses for the calculation.

The only data point for which U.S. News appears to provide data for all ranked schools is test scores.

Upon closer inspection, though, the data provided is not the data used in the rankings.

A scatterplot graph showing that 45 schools have a graduate indebtedness rank with no data provided.

There are 45 schools (21% of the 215 schools with a Graduate Indebtedness Rank) for which the data is noted by U.S. News as "N/A."

How can U.S. News produce a Graduate Indebtedness Rank for these schools if the data is Not Available or Not Applicable?

There must be data used. Why is it not disclosed?


The rankings have suspicious trends.

There is overlap between numerous data points used by U.S. News and those contained in the IPEDS database, but there are discrepancies between the databases.

Institutions self-report their data to both U.S. News (if they choose to do so) and to IPEDS. Schools have an incentive to make their data appear as positive as possible when reporting to U.S. News, and there is evidence that schools may manipulate their data to improve their rankings (Morson and Schapiro, 2017).

On the other hand, by law, schools must accurately complete the IPEDS survey, which is tied to their ability to participate in federal funding programs.

The chart at the right shows the IPEDS and U.S. News data points for Full Time Faculty. Both follow U.S. News' formula.

While there is overlap for some schools, there are considerable variances for others (as identified by the longer bars between the two points).

A scatterplot graph showing no apparent relationship between the percentage of classes with under 20 students and U.S. News ranking.

If the data and calculations provided by U.S. News were fully transparent and accurate, one would expect to see a linear (or other distinguishable) relationship between the factors and the rankings.

The percentage of classes with under 20 students is plotted against the U.S. News rankings but does not appear to show any distinguishable relationship. This is one of the factors for which U.S. News does not report the weighting nor provide all of the data used for the rankings.

With the extensive documentation provided on how U.S. News calculates its Best Colleges rankings, and the detailed rankings pages provided in their College Compass database, U.S. News appears to be transparent.

However, as documented above, there are gaps in both the factor weightings and the actual data points, that make it impossible to reproduce the U.S. News rankings with the data and formulas U.S. News provides.

This chart shows the results of taking U.S. News' sub-rankings and multiplying it by the weighting for each sub-ranking, and dividing this result by the total weighting for available sub-rankings. This last step adjusts for data not made available by U.S. News.

U.S. News (and the calculations used for this chart) takes the highest value for all schools (Williams College in 2022), and then the scores calculated above are divided by the top school's score. This is similar to when teachers curve a test.

There are some schools for which the calculated score is quite close to the U.S. News score (the top three schools, for example) but others are meaningfully different (the longer the horizontal line, the bigger the difference).

The biggest differences seem to be among the lower-ranked schools, and generally the calculated scores are higher for higher-ranked schools, and lower for lower-ranked schools. The U.S. News scores seem to be less extreme.

Scientific research requires reproducibility and the U.S. News rankings cannot be reproduced. If they are simple calculations as described, it should be simple to reproduce the rankings.

A group of students at Reed College used "complicated statistics" to attempt to reproduce the U.S. News rankings, and were successful, though they found discrepancies in what U.S. News reported for their school, which chooses not to provide data to U.S. News (Lydgate, 2019).

A stacked area chart showing the frequent changes to the ranking factors and their weightings.

Lastly, the ranking factors and weights change continuously. This may be because U.S. News seeks to better reflect what people care about (such as the recent inclusion of social mobility and graduate indebtedness), or it may be a way to cause movement in the rankings that allow for "newsworthy changes."

When the rankings were originally published, they were exclusively based on reputation. Quantitative measures were first introduced in 1987 (Grewal, R., Dearden, J. A., & Llilien, G. L. (2008).


Is more always better than less? It depends what the "more" is.

All ranking systems are inherently biased; by definition, they must incorporate judgment about what matters, and what matters most. As users we need to decide whether we agree with those judgment calls and should consider holding organizations, like U.S. News, accountable for providing transparency behind their data.

U.S. News has decided that reputation and financials matter most, though it is all reported under the guise of academic quality.

When asked what is the basis for the assumption that the more money a school has, the better quality of the education, Robert Morse, chief data strategist for U.S. News and the man in charge of the rankings, replied:

An illustration of Robert Morse, with a quote arguing that more spending means a richer academic experience.
0:00

Robert Morse about why U.S. News includes financially oriented factors in its best college rankings.

How to do better

U.S. News can increase transparency and do more diligence to ensure accuracy and reliability of self-reported data.

Colleges can also refocus their efforts and resources on fulfilling their missions and serving students rather than prioritizing the rankings.

Perhaps most importantly, students and parents can start with a blank sheet of paper and think about what is important in a college experience. Armed with a list of priorities, they can find reliable data to help them narrow down the list of schools. One such source is the government survey data available through IPEDS.

There are also other ranking systems that may be better aligned with student and parent priorities or are more direct about what is being measured:

  • Academic Influence ranks schools based on their belief that influential graduates are the best marker of a quality education.
  • Forbes believes that "schools that do an especially good job of reaching out to lower-income students, kids denied the benefits of first-rate suburban or private high schools, summer camps, tutors, etc., should be recognized when they promote economic and social mobility" (Vedder, 2021) and incorporates this into their ranking system, which is "based on outcomes, not reputation and test scores."
  • Niche is a ranking system that includes student feedback alongside the quantitative data (from sources such as the U.S. Department of Education).