How Azimuth Measures College Value

Azimuth provides a single, clear answer to a complex question: What do students actually get from attending a college?

Our rankings combine data on earnings, affordability, access, and mobility to help students and families understand the real value of a college education.

We use only publicly available data, primarily from the U.S. Department of Education's College Scorecard (April 2025 release) and the Center for American Progress loan repayment dataset. All results apply to undergraduate outcomes at public and nonprofit four-year institutions.

Azimuth's rankings are built from four pillars. Together they capture both who a college serves and how students fare after they leave.

Each institution receives a score in all four areas, which are then combined into an overall ranking.

What it measures:

How well a college pays off financially in the long run.

ROI combines three ideas:

• Long-term earnings: What graduates earn 6, 8, and 10 years after enrolling. We prioritize 10-year earnings because they best reflect career trajectory.
• Earnings beyond expectations: How much more (or less) graduates earn compared with similar students at other colleges. This isolates the college's contribution to economic outcomes.
• Loan repayment patterns: How quickly borrowers pay down their federal student loans after graduation.

Why it matters:

Two colleges may have similar tuition, but dramatically different long-term payoffs. ROI highlights colleges where graduates see strong earnings and positive financial momentum.

Where the data comes from:

U.S. Department of Education College Scorecard (earnings + repayment), April 2025 release.

What it measures:

How much students actually pay — not sticker price.

Affordability looks at:

• Net price by income level: What low-, middle-, and high-income families typically pay after grants and scholarships.
• Student debt: How much students borrow, including higher- and lower-debt scenarios.
• Parent PLUS borrowing: Debt taken on by parents.

Why it matters:

Families need a clear view of what college will cost them, not just a published number.

Where the data comes from:

U.S. Department of Education College Scorecard (net price, federal loan data), April 2025 release.

What it measures:

Whether a college offers meaningful educational opportunities to a broad range of students.

Access includes:

• Pell Grant enrollment: Share of students from low-income families.
• First-generation students.
• Admission patterns, when data is available.
• Scale: How many Pell students the college actually serves.
• Institutional risk: Whether the college serves many Pell students despite limited financial resources (measured using endowment per Pell student).

Why it matters:

Two colleges might have similar outcomes, but one may be serving far more low-income or first-generation students than the other.

Where the data comes from:

U.S. Department of Education College Scorecard, April 2025 release.

What it measures:

Whether low-income students actually experience upward mobility after attending the college.

Mobility includes:

• Earnings for low-income students 10 years after enrolling.
• Earnings compared with peers at similar institutions.
• Pell graduation rates, adjusted for peer groups.
• Aggregate mobility impact, combining outcomes and the number of low-income students served.
• A mobility “quadrant” that shows whether a college is a
  • • Mobility Engine (high access + strong outcomes),
  • • Selective Achiever,
  • • Opportunity Builder, or
  • • Under-Resourced Institution.

Why it matters:

Families care about whether students from similar backgrounds thrive and move up economically.

Where the data comes from:

U.S. Department of Education College Scorecard, April 2025 release.

Our default ranking balances long-term payoff with broad opportunity:

• 30% Mobility
• 30% Return on Investment
• 20% Access
• 20% Affordability

This produces a picture of “overall value” that rewards both economic outcomes and inclusive opportunity.

You can also explore alternative lenses on our site:

• Outcome Lens (60% Return)
• Affordability Lens (60% Affordability)
• Mobility Lens (70% Mobility)

When data is missing:

• We use fallbacks (e.g., 10-year → 8-year → 6-year earnings).
• If a metric is still missing, we use a median value so colleges aren't penalized simply because federal data wasn't reported.
• No institution is excluded from the rankings unless it fails basic coverage rules.

We do not guess or estimate outcomes beyond the published data.

For this version, Azimuth includes only public and private nonprofit four-year institutions.

We excluded for-profits and 2-year colleges because:

• Data coverage differs significantly,
• Their missions and student populations are meaningfully different, and
• Many of the long-term outcome measures are not directly comparable.

We plan to revisit expanded categories in future editions.

Azimuth uses publicly available federal data as the foundation for all rankings. We perform our own analysis on this base data to calculate proprietary metrics — we do not modify the underlying government data.

U.S. Department of Education — College Scorecard

What it includes:

• Earnings at 6, 8, and 10 years after enrollment
• Net price by income level
• Pell Grant and first-generation student shares
• Federal loan and repayment metrics
• Graduation and completion rates
• Program-level earnings (by CIP code)

Center for American Progress — Loan Repayment Trajectory Dataset

This tracks a cohort of federal borrowers for five years after entering repayment — a far longer view than standard 3-year default rates.

The model is re-run as new federal data becomes available.

Azimuth updates all scores together to ensure internal consistency across pillars.

We welcome feedback from researchers, policymakers, families, and institutions — each iteration improves the clarity and fairness of the model.

A deeper technical appendix — with details on models, z-scoring, fallbacks, and peer comparisons — is available on our Technical Methodology page.