For decades, college rankings have run on the same inputs: reputation surveys, acceptance rates, alumni giving, and institutional prestige. They reward schools for being selective and famous — not for what they actually do for the students who enroll. We think that answers the wrong question.
So we ask a different one: what happens to students after they graduate?
That question has an answer. Researchers at Opportunity Insights, led by economist Raj Chetty, followed roughly 30 million students through anonymized tax records to measure where they started and where they ended up. Paired with the U.S. Department of Education's College Scorecard, it lets us measure the things that actually matter:
- Earnings — what graduates actually make, ten years after they enroll.
- Social mobility — whether a college lifts low-income students up the income ladder, or simply admits students who were already at the top.
- Debt — what students borrow, measured against what they go on to earn.
- Completion — whether students actually finish what they start.
We are not grading institutions in isolation. We are measuring transformation — the distance a college moves the students it serves. A school that takes students from the bottom and launches them into the top is doing something that a school which admits the already-wealthy, and takes credit for their success, is not.
And we don't ask you to take it on faith. Every weight, every formula, and every data source is published below — and any school can raise its score simply by reporting more of its data.
The Core Algorithm
Every school receives a composite score from 0–100 built from four weighted pillars. Missing data triggers a transparency penalty that reduces the final score. Here's the exact math:
Formula 1: Normalization
Formula 2: Weighted Composite
Formula 3: Final Score with Transparency Penalty
Interactive Score Simulator
Adjust the sliders to see how a hypothetical school would score across all four pillars.
Composite Score
After transparency penalty: 65.2
Pillar 1 — Academic Quality 15% of composite
| Metric | Weight | Range | Inverted | Source |
|---|---|---|---|---|
| Completion rate | 25% | 0 → 1 (100%) | No | IPEDS |
| Retention rate | 20% | 0 → 1 (100%) | No | IPEDS |
| Full-time faculty rate | 15% | 0 → 1 (100%) | No | IPEDS |
| Diversity index | 10% | 0 → 0.80 | No | IPEDS |
| THE Research Score | 15% | 0 → 100 | No | THE |
| THE Teaching Score | 10% | 0 → 100 | No | THE |
| THE Citations Score | 5% | 0 → 100 | No | THE |
Requires at least 25% weight coverage to produce a score. THE data only available for ~150 research universities; schools without THE data automatically reweight to the remaining metrics.
Pillar 2 — Economic Outcomes 30% of composite
| Metric | Weight | Range | Inverted | Source |
|---|---|---|---|---|
| Earnings at 10 years | 30% | $15,000 → $120,000 | No | Scorecard |
| Debt-to-earnings ratio | 25% | 0 → 2.0 | Yes | Computed |
| Loan repayment rate (3yr) | 20% | 0 → 1 (100%) | No | Scorecard |
| % earning more than HS grads | 15% | 0 → 1 (100%) | No | Scorecard |
| Median debt | 10% | $0 → $40,000 | Yes | Scorecard |
This pillar carries the highest weight (30%) because economic advancement is college's primary promise. Schools where graduates earn more, repay debt faster, and carry less debt relative to income score highest.
Pillar 3 — Social Mobility & Capital 35% of composite
This is what makes CollegeRanker different. No other ranking system measures whether a college actually propels low-income students into the upper class — because until Chetty's 2017 study, the data didn't exist. Now it does, and we weight it most heavily.
| Metric | Weight | Range | Inverted | Source |
|---|---|---|---|---|
| Mobility rate (Q1 → Q5) | 25% | 0 → 0.03 (3%) | No | Chetty MRC |
| Success rate (bottom 20%) | 15% | 0 → 0.70 (70%) | No | Chetty MRC |
| Economic connectedness | 15% | 0 → 2.0 | No | SC Atlas |
| Bottom-20% access | 10% | 0 → 0.25 (25%) | No | Chetty MRC |
| Friending bias | 10% | −0.2 → 0.4 | Yes | SC Atlas |
| Pell Grant rate | 10% | 0 → 0.80 (80%) | No | Scorecard |
| Income diversity | 10% | 0 → 1.0 | No | Chetty OI |
| Volunteering rate | 5% | 0 → 0.35 (35%) | No | SC Atlas |
How We Measure Mobility
-
Mobility Rate
P(child in top 20% | parent in bottom 20%)
The probability that a student from the bottom income quintile reaches the top quintile as an adult. This is the single most powerful measure of whether a college delivers on the American Dream.
-
Success Rate
P(top 20% | bottom 20%)
Among students who come from the bottom 20% of families, what share reaches the top 20%? This differs from mobility rate by not requiring a specific income quintile destination.
-
Economic Connectedness
EC = P(high-SES friend) / P(low-SES friend)
Measures whether students form cross-class friendships. Chetty's research found this is the strongest predictor of upward mobility — more than school quality, more than family structure.
All mobility and social capital data is from Opportunity Insights (Chetty et al., 2017–2022), covering 30+ million anonymized tax records. No other ranking system uses this data.
Pillar 4 — Value & Affordability 20% of composite
| Metric | Weight | Range | Inverted | Source |
|---|---|---|---|---|
| Average net price | 30% | $0 → $50,000 | Yes | Scorecard |
| Net price ($0–30K families) | 25% | $0 → $30,000 | Yes | Scorecard |
| Federal loan rate | 20% | 0 → 1 (100%) | Yes | Scorecard |
| Median debt | 15% | $0 → $40,000 | Yes | Scorecard |
| Debt-to-earnings ratio | 10% | 0 → 2.0 | Yes | Computed |
Unlike other rankings that use sticker price, we use net price (tuition minus grants/scholarships) and break it out by family income bracket. A $60K sticker price means very different things to different families.
Transparency Penalty
Schools that report data comprehensively keep their full score. Schools that withhold data get penalized — because if an institution won't show you its outcomes, that tells you something.
Data Completeness Scoring
- Critical fields (×3 each)
- Completion, Earnings, Net Price, Enrollment, Retention
- Important fields (×2 each)
- Acceptance, Debt, Tuition, SAT/ACT, Repayment, Pell, D/E, Net price by income
- Valuable fields (×1 each)
- Faculty rate, Faculty salary, Diversity, Earnings 6yr, Loan rate, First-gen, Part-time
- Bonus fields (×0.5 each)
- Connectedness, Friending bias, THE scores, Mobility rate, Success rate, Innovation
Penalty Function
final = raw × (0.50 + 0.50 × completeness)
- 100% complete
- no penalty
- 80% complete
- 10% penalty
- 60% complete
- 20% penalty
- 40% complete
- 30% penalty
- 20% complete
- 40% penalty
- 0% complete
- 50% penalty
Income Trajectories
The most important thing a college can do is change a family's economic trajectory across generations. Using Chetty's Mobility Report Card data, we track what happens to students from the bottom 20% of the income distribution — and compare it to what the top 20% already have.
Generational Mobility
Each line shows how a student from the bottom income quintile fares after attending a given type of college.
-
High-Mobility Schools
Institutions like SUNY Stony Brook, Cal State LA, and UT Dallas take bottom-20% students and give them a 15-25% chance of reaching the top 20%. Their graduates earn more than 80% of the population within 15 years.
-
Average Schools
Most colleges provide moderate upward mobility. A bottom-20% student has roughly a 5-10% chance of reaching the top 20% — better than the 2-3% chance without college, but far from transformative.
-
Low-Mobility Schools
Some institutions — including a surprising number of prestigious names — serve almost exclusively wealthy students and produce graduates who end up roughly where they started. Their mobility rates are barely above the no-college baseline.
Ranking-Specific Weighting
Our flagship "Best Colleges" ranking uses the standard composite above. But different questions require different weights. Here's how each ranking adjusts the formula:
| Ranking | Academic | Economic | Mobility | Value | Other |
|---|---|---|---|---|---|
| Best Colleges (Flagship) | 20% | 30% | 25% | 25% | — |
| Best Value | 15% | 35% | 25% | 25% | — |
| Social Mobility | 0% | 30% | 55% | 15% | — |
| Social Capital | 10% | 0% | 0% | 0% | 90% SC Atlas |
| Highest Earnings | 0% | 100% | 0% | 0% | Direct raw earnings |
| Most Affordable | 0% | 0% | 0% | 100% | Inverse net price |
| Most Selective | 0% | 0% | 0% | 0% | Inverse acceptance rate |
| Best Completion | 0% | 0% | 0% | 0% | Raw completion rate |
| Safest (Financial) | 0% | 0% | 0% | 0% | Repayment + Completion + Debt |
How We Compare to Other Rankings
| What's Measured | CollegeRanker | U.S. News | Forbes | Niche |
|---|---|---|---|---|
| Social mobility data (Chetty) | ✓ | × | × | × |
| Social capital / cross-class networks | ✓ | × | × | × |
| Innovation / patent inventor rate | ✓ | × | × | × |
| Net price by family income bracket | ✓ | × | × | × |
| Earnings at multiple time horizons | 6/8/10yr | Partial | 10yr only | × |
| Debt-to-earnings ratio | ✓ | × | × | × |
| Loan repayment health | ✓ | × | × | × |
| Transparency / data completeness penalty | ✓ | × | × | × |
| Peer reputation surveys | None (0%) | 20% of score | Partial | Student reviews |
| Subjective inputs | Zero | Multiple | Multiple | Majority |
| All data publicly verifiable | ✓ | Partial | Partial | × |
Our Data Sources
Every data point in our algorithm comes from one of these seven sources. All are publicly accessible. None require our interpretation — we apply the data; we don't create it.
-
U.S. Department of Education — College Scorecard
~6,400 institutions. Tuition, earnings, debt, completion rates, admissions, financial aid, program enrollment. Updated annually.
collegescorecard.ed.gov/ → -
Opportunity Insights — Mobility Report Card
2,202 schools. Mobility rates, parent and child income distributions, access trends. Built from 30M+ anonymized tax records.
opportunityinsights.org/data/ → -
Opportunity Insights — Social Capital Atlas
2,586 schools. Economic connectedness, friending bias, volunteering rates, community support.
opportunityinsights.org/data/ → -
Opportunity Insights — Innovation & Patent Data
423 schools. Patent inventor rates by income quintile, citation counts.
opportunityinsights.org/data/ → -
NCES IPEDS
Integrated Postsecondary Education Data System. Enrollment, demographics, faculty, institutional characteristics.
nces.ed.gov/ipeds/ → -
Times Higher Education World Rankings
~150 U.S. schools. Research quality, teaching scores, citation impact.
www.timeshighereducation.com/world-university-rankings → -
Bureau of Labor Statistics
Occupational salary data, employment projections, and industry growth rates.
www.bls.gov/ooh/ → -
Common Data Set (via collegedata.fyi)
~698 schools. School-published admissions detail: enrollment yield, real SAT/ACT submission rates, Early Decision/Action splits, and waitlist activity. Each figure links back to the institution’s own CDS filing.
www.collegedata.fyi/ → -
College Scorecard — Field of Study
Program-level (CIP) early-career median earnings by major and credential level, for all fields. Powers the by-major earnings in the Students Like You simulator, shown distinctly from school-wide 10-year medians. Earnings are suppressed for small programs.
collegescorecard.ed.gov/data/ → -
BLS Consumer Price Index (CPI-U)
Used to express dollar figures from different vintages in today’s dollars. Opportunity Insights mobility incomes are reported in 2015 dollars and College Scorecard earnings are measured a few years before each release; we CPI-adjust both to the current year so cost, earnings, and income never mix vintages. Rates and shares are not dollar figures and are left unadjusted.
www.bls.gov/cpi/ →
Why This Matters
Choosing a college is one of the most consequential financial decisions an American family will ever make. The average student borrows over $30,000. Families sacrifice savings, take second jobs, remortgage homes. They do this because they believe — correctly — that higher education is the most reliable path to a better life.
But not all colleges deliver on that promise equally. Some institutions take students from poverty and launch them into prosperity at rates that seem almost miraculous. Others charge six figures and produce graduates who struggle to repay their loans. The difference between these outcomes is measurable, and it's in the data.
CollegeRanker exists to make that data visible, accessible, and actionable. Not because an administrator said so in a survey. Because the tax records prove it.
