← Index | Demand Side — Colleges & Admissions › Enrollment Management: How Colleges Hit Yield Targets

College Enrollment Management: How Colleges Hit Yield Targets

Source: college_enrollment_management.md

College Enrollment Management: How Colleges Hit Yield Targets

Research compiled from NACAC reports, institutional Common Data Sets, and enrollment management literature.

Admit Number Setting Methodology

The Core Formula

Colleges determine how many students to admit using a straightforward but high-stakes calculation:

Admits_needed = Target_class_size / Expected_yield_rate

For example, if a college wants 1,700 freshmen and expects a 70% yield, it admits ~2,430 students. If yield is predicted at 40%, it must admit ~4,250 — a dramatically different risk profile.

Yield Prediction Models

Institutions use multi-factor predictive models built on:

  1. Historical yield data — typically a 5-year rolling average as the baseline, weighted toward recent years
  2. Demographic segmentation — yield is predicted per subgroup (in-state vs. out-of-state, legacy vs. non-legacy, financial aid recipients vs. full-pay, recruited athletes vs. general applicants)
  3. Behavioral engagement signals — campus visit attendance, email opens, portal logins, webinar participation, interview completion
  4. Cross-admit competitor modeling — estimating which peer schools an applicant is likely to choose over yours (e.g., Harvard vs. Yale cross-admits historically split roughly 60/40 Harvard)
  5. Financial aid package competitiveness — how the offered package compares to likely competitor offers

Third-party platforms (Encoura, EAB, Capture Higher Ed) provide enrollment prediction models that incorporate behavioral data, academic profiles, and geographic data to score individual applicants' likelihood of enrolling.

Yield Rates at Elite Schools (Class of 2029)

School Yield Rate Acceptance Rate
Harvard 84% ~3.4%
MIT ~85% ~3.9%
Stanford ~81% ~3.9%
Princeton 78.3% ~5.7%
Yale 67.7% ~5.7%
Columbia 67.1% ~5.5%
Brown 67.3% ~5.0%
UPenn 67.9% ~5.7%
Cornell 68.4% ~7.9%
Dartmouth 63.7% ~6.2%

Key insight for simulation: HYPSM yield rates cluster at 78-85%, meaning these schools need to admit only ~1.18-1.28x their target class size. Lower-tier selective schools with 30-50% yield must admit 2-3x their target, introducing far more uncertainty.

Overbooking and Safety Margins

Colleges intentionally overshoot their target slightly (typically 2-5% above target enrollment), because of summer melt — the phenomenon where deposited students fail to matriculate. Melt rates vary:

Georgia Tech, for example, reports melt of ~2% for in-state, ~8% for out-of-state, and ~15% for international students. UC Irvine famously had to rescind admission offers when 850 more students than expected accepted their offers.

ED/EA Class Fill Percentages

How Early Rounds Fill the Class

Early Decision (binding) and Early Action (non-binding) serve fundamentally different enrollment management purposes:

Percentage of Class Filled by Early Rounds

Schools with ED programs (binding commitment):

School % Class via ED Notes
Northwestern 53% ED I only
UPenn 53% ED I only
Duke 51% ED I only
Dartmouth 48% ED I only
WashU ~60% ED I + ED II
NYU ~60% ED I + ED II
Middlebury 68% ED I + ED II
Grinnell 65% ED I + ED II
Pomona 55% ED I + ED II
Brown 40% ED I only
Cornell 40%+ ED I only
Boston University 44% ED I + ED II
UVA 31% ED I only

Trend: The share of students enrolled through ED rose from 38% to 54% on average across 66 selective colleges between 2015/16 and 2024/25. Schools are increasingly relying on binding early rounds.

HYPSM schools (REA/EA — non-binding):

Harvard, Yale, Princeton, Stanford, and MIT use Restrictive Early Action (REA) or Early Action (EA), which is non-binding. These schools do not lock in students through early rounds the same way ED schools do:

Key simulation parameter: ED schools fill 40-60%+ of their class before RD; REA/EA schools fill 15-25% early but with lower commitment certainty.

ED II (Second Binding Round)

ED II (January deadline) allows schools to fill remaining gaps after ED I:

Financial Aid as Enrollment Lever

Need-Blind vs. Need-Aware

Need-blind institutions (do not consider ability to pay in admissions):

Need-aware institutions (consider ability to pay):

Merit Scholarships as Enrollment Tools

Schools that DO NOT offer merit aid (need-based only):

Schools that use merit aid strategically:

Financial Aid Impact on Yield

Student Category Yield Impact
Full-pay students Lower yield (more options, price-sensitive to alternatives)
Full-need-met students at top schools Very high yield (best package they'll see)
Merit award recipients Higher yield than non-recipients at same school
Students with competing merit offers Yield depends on relative package size

Key simulation insight: Need-blind HYPSM schools don't use financial aid as an enrollment lever. Schools ranked 15-50 use merit aid as a primary yield management tool. The simulation should model merit aid as a yield multiplier for non-HYPSM schools.

Waitlist Activation Model

Why Waitlists Exist

Waitlists serve as a safety valve for yield uncertainty. Rather than admitting extra students and risking over-enrollment, colleges can:

  1. Admit a conservative number in RD
  2. Wait until May 1 deposit deadline results come in
  3. Fill remaining gaps from the waitlist

When Waitlists Activate

Primary trigger: Actual deposits fall short of target enrollment after the May 1 National Candidates Reply Date.

Typical timeline:

Decision factors:

Waitlist Statistics at Ivy League Schools

School Avg. Waitlist Accept Rate Range Notes
Harvard Not disclosed — Rarely uses waitlist
Princeton ~5% 0.15%-16.4% Used waitlist in ~2/3 of recent cycles
Cornell ~4.2% varies 388 admitted from waitlist for Class of 2028
Dartmouth ~4.1% varies 21-year average
UPenn ~2-6% 0.5%-17% Highly variable year to year
Columbia Low single digits varies Unpredictable

Key patterns:

Waitlist Size Strategy

Schools typically place 3-10x more students on the waitlist than they expect to admit from it, because:

Need-Blind vs. Need-Aware: Enrollment Management Differences

Need-Blind Enrollment Management

Schools operating need-blind face unique challenges:

  1. Revenue uncertainty: Cannot predict the financial aid burden of the incoming class until admits decide
  2. No price lever: Cannot adjust admit offers based on ability to pay
  3. Yield management via prestige: Must rely entirely on institutional reputation, campus experience, and program quality
  4. Endowment dependence: Financial aid budget drawn from endowment returns (typically 5% annual draw rate)

Practical reality: Only schools with endowments exceeding ~$3B per student (Harvard, Yale, Princeton, Stanford, MIT) can sustainably maintain need-blind + meet-full-need policies without it affecting institutional finances.

Need-Aware Enrollment Management

Most selective schools (even many in the top 30) are need-aware, which enables:

  1. Revenue modeling: Predict net tuition revenue from each admitted cohort
  2. Strategic merit aid: Offer discounts to high-desirability students to improve yield
  3. Admit class shaping: Ensure the class generates enough revenue to fund financial aid commitments
  4. Marginal admits: For borderline applicants, ability to pay can tip the decision — not the primary factor, but a tiebreaker

Key distinction: Need-aware status affects primarily borderline decisions. Clearly admissible students are rarely affected. The impact is concentrated on the last 5-15% of the admit pool.

Simulation Algorithm Recommendations

Based on this research, here are the key enrollment management parameters for the simulation:

1. Admit Number Calculation

function calculateAdmits(college) {
  const targetSize = college.targetClassSize;
  const expectedYield = college.historicalYield; // segmented by round
  const meltBuffer = 1.02; // 2% overshoot for melt

  // ED round (binding)
  const edTarget = targetSize * college.edFillPercent;
  const edAdmits = edTarget; // ~100% yield for ED

  // EA/REA round (non-binding)
  const eaTarget = targetSize * college.eaFillPercent;
  const eaAdmits = eaTarget / college.eaYield;

  // RD round (fill remainder)
  const rdTarget = (targetSize - edTarget - eaTarget) * meltBuffer;
  const rdAdmits = rdTarget / college.rdYield;

  return { edAdmits, eaAdmits, rdAdmits };
}
Parameter HYPSM Ivy+ Near-Ivy Selective
Overall yield 78-85% 63-68% 40-55% 25-40%
ED fill % N/A (REA) 40-53% 35-50% 25-40%
EA/REA fill % 15-25% N/A N/A N/A
ED yield N/A ~98% ~95% ~90%
EA yield 70-80% N/A N/A N/A
RD yield 60-70% 30-45% 20-35% 15-25%
Waitlist use Rare 2-6% 5-15% 10-25%
Melt rate 1-2% 2-3% 3-5% 5-10%
Merit aid? No Some Yes Yes

3. Waitlist Activation Logic

function activateWaitlist(college, depositsReceived) {
  const gap = college.targetClassSize - depositsReceived;
  if (gap <= 0) return []; // no waitlist needed

  // Admit from waitlist with predicted acceptance rate
  const waitlistOffers = gap / college.waitlistAcceptRate;
  // Sort waitlist by score, fill specific profile gaps
  return selectFromWaitlist(college.waitlist, waitlistOffers);
}

4. Financial Aid Yield Modifier

function aidYieldModifier(college, student) {
  if (college.needBlind && college.meetsFullNeed) {
    return 1.0; // no yield effect from aid at top schools
  }
  if (student.meritAward > 0) {
    return 1.0 + (student.meritAward / student.totalCost) * 0.3;
    // merit award increases yield probability
  }
  if (student.needGap > 0) {
    return 1.0 - (student.needGap / student.totalCost) * 0.5;
    // unmet need decreases yield probability
  }
  return 1.0;
}

5. Round-by-Round Enrollment Flow

Round 1 (ED):    Binding admits → ~98% yield → fills 40-60% of class
Round 2 (EA/REA): Non-binding admits → 60-80% yield → fills 15-25% (HYPSM only)
Round 3 (ED II):  Binding admits → ~95% yield → fills 5-15% additional
Round 4 (RD):     Regular admits → 20-50% yield → fills remainder
Round 5 (May 1):  Deposit deadline → tally actual enrollment
Round 6 (WL):     Waitlist activation → fills gaps (0-15% of class)
Round 7 (Summer): Melt management → replace lost deposits

Sources