Outward Bound USA · Open Enrollment · Executive Director and Board Briefing

Why OE enrollment fell, and how we stabilize it in 2027 and 2028

We tested our 2026 planning research against outside market data. Where the two agree, the finding stands; where they disagree, we say so. Prepared November 18, 2026 for 2027 planning. 2026 figures are provisional.

Bottom line

Families still want what we offer. Fewer of them are choosing us, and the ones who do are taking longer to commit.

OE has fallen from 9,020 students in 1995 to about 2,114 in 2026. Outside data and our own records agree on three causes:

  1. Our share of the market collapsed, not the market itself. Outdoor participation is at a record 183 million Americans. The number of teen households that can afford a course roughly tripled. Our share of those households fell about twelvefold. A field of direct competitors, almost all founded between 1983 and 2009, grew up around us, and families started researching through search, social media and now AI.
  2. 2026 conditions make families slow to commit. Consumer confidence is in a long slide, savings are thin, and 73% of parents say inflation changed their summer plans. We see the same thing in our funnel: applications arrive later (98 days before start, down from 115), consultations rose almost sevenfold, and 16% of applicants ask for aid.
  3. We have less to sell. Listed courses fell from 422 in 2014 to 243 in 2026, and one in five 2026 courses was cancelled. Courses got shorter, and long programs shrank. Where a School added viable departures (HIOBS: 26% more running courses), students rose 22% and fill held at about 93%.

What follows for 2027–28: the market is not about to rescue us. The teen population is flat, and high school graduates peaked in 2025. Broad price cuts and generic awareness campaigns don't address any of the three causes. Stability will come from the parts we control: run more of the courses that work, cancel fewer for avoidable reasons, make committing feel safer and simpler, and measure enrollments rather than leads. The arithmetic in Section 7 shows that these levers are large enough to absorb a weak market year.

−77%OE students, 1995 → 2026 (9,020 → ~2,114). Down about 50% since 2019 even with Veterans removed. Internal
183.2MAmericans who participated in outdoor recreation in 2025, a record. Kids 6–12 up 5%. External
0.82% → 0.065%OE's share of affordable teen households, 1995 → 2026. A roughly 12× collapse while that pool tripled. Internal
115 → 98 daysMedian time from application to course start, 2024 → 2026. Travel advisors report the same shortening of booking windows. Both agree
422 → 243Listed OE courses, 2014 → 2026. In 2026, 49 were cancelled and 194 ran, at 87% fill. Internal
+26% → +22%HIOBS ran 26% more courses and enrolled 22% more students, with fill steady near 93%. More viable courses meant more students. Internal
1 · The evidence test

Each explanation, checked against outside data and our own

Boards hear many explanations for enrollment decline. This table checks each one against two independent sources: published market data External and our own records Internal. A finding is confirmed only when both point the same way.

ExplanationWhat outside data saysWhat our data saysVerdict
“Families don't want outdoor programs anymore” A record 183.2M Americans took part outdoors in 2025, and kids 6–12 rose 5%.E1 Parents of 24.6M children want a structured summer program, but only about 12M are enrolled.E2 Courses that run still fill at 87%. When HIOBS added viable departures, students followed.I1 Not supportedThe demand is there. Turning it into a booked course is the hard part.
“Competitors are taking our families” The Gap Year Association directory lists 91 providers.E16 NOLS now markets 20+ gap programs with credit, certifications and 529/federal-aid pathways.E17 The outdoor economy is $697B.E19 Of the 26 direct competitors we track, 17 were operating by 1995. Nine more launched 1995–2009, when OE fell fastest. None has launched since.E20 Our share of affordable teen households fell ~12×.I1 ConfirmedThe field formed between 1983 and 2009. Since then the pressure comes from those programs growing and being easy to compare online, not from new entrants. We can't trace individual families who switched.
“We're too expensive” CPI-U roughly doubled since 1995.E8 73% of parents say inflation changed summer plans, and half of those cut camps.E3 58% of adults say prices worsened their finances.E7 Our list price rose 68% since 1995, less than inflation. In 2023–26, a 10% price rise moved fill by about 0 points, and cheaper courses did not fill better.I2 But price per day rose 62% since 2018, and 16% of applicants request aid. PartlyCost causes real hesitation, but it didn't cause the decline. Total trip cost and fewer days for the money matter more than the sticker price.
“The economy is making families hesitate” The Conference Board index averaged ~104 in 2024, ~96 in 2025 and ~90 in 2026.E4 Michigan sentiment hit a record low of 44.8 in May 2026.E5 Saving rate: 7.3% (2019) → ~4.7% (2026).E6 Advisors citing 1–3 month booking windows: 9.8% → 25.4%.E9 Median days from application to start fell from 115 to 98. Consultations at the 2027 launch rose from 8 to 55. Early summer applications are 69, down from 98 in fall 2024.I1 ConfirmedThis pressure comes and goes with the economy. It hits at the moment of commitment, which is exactly where we can act.
“Marketing isn't reaching people” When Google shows an AI summary, searchers click a result half as often (8% vs 15%).E13 Apple's tracking opt-in (2021) made ad targeting and measurement harder.E15 USA Gap Year Fairs fell from 43 to 16.I1 Most lead records can't be linked to later applications. 138 people started a 2027 application and stopped; only 1 has a lead record. Their courses carry about $750K in tuition.I3 ConfirmedReaching families is harder for everyone. Losing track of applicants who were already interested is our own gap.
“We cut capacity and changed the mix” Specific to OB; no outside benchmark applies. Outside shocks add to it, such as July 2026 wildfire smoke that affected ~120M people.I1 Listed courses: 422 → 243. Adult listings: 151 → 30. Days per student: 16.9 → 13.4. NWOBS closed in 2024. Intercept fell 88%.I1 Internal, controllableOur largest single lever, and the one with the clearest evidence.
“Cancellations are getting worse” No outside benchmark applies. Cancellation rates: 24.1% (2013–16), 15.6% (2017–19), 21.3% (2024–25), 20.2% (2026). These swing from year to year without a steady trend. 9 of the 49 cancellations in 2026 were wildfire-related.I1 Not supportedCancellations are significant, but they aren't steadily rising. Reason codes are too weak to tell what share is low demand.
“Fewer teenagers” Added here U.S. high school graduates peaked in 2025 (~3.9M) and decline from here.E10 Children born in 2011–2014, who are 13–16 in 2027, number about 3.95M a year. The 2007 peak was 4.32M.E11 Our best-performing products serve ages 13–14 and 14–16. Programs for 16–18 and gap-year students depend more on the specific trip.I1 Emerging headwindThis didn't cause the past decline. It does mean population won't lift us in 2027–28, especially for gap-year programs.
“Our families can't afford it now” Added here High-income households now account for an outsized share of U.S. spending: ~46% for the top 10% (Moody's) versus ~23% (BLS).E12 45% of high-income children attend summer camp, compared with 13% of low-income children.E2 Still, 53% of $100K+ earners say prices hurt their finances.E7 Average total cost ($5,052) roughly kept pace with inflation since 2018. 16.4% of applicants request aid, with an average award of $2,287.I1 Mostly confirmedOur core market can still pay but is more cautious. That calls for reassurance and clear aid information, not discounts across the board.
Structural · years in the making

Share of choice

More providers, more substitutes and fragmented research channels. Brand recognition still helps, but it no longer guarantees we're on the shortlist.

Control: partial. We can win more of the comparisons families make, but we can't reverse the trend.
Cyclical · 2024–2027

Commitment friction

Low confidence, thin savings, shorter booking windows and travel risk. Families keep their options open: they delay, compare and ask for help.

Control: high, at the moment of purchase: total cost, deposits, aid and consultation.
Our choices · reversible

Capacity and mix

Fewer listed courses, avoidable cancellations, shorter courses and fewer long programs. These reduced enrollment regardless of the market.

Control: highest. The evidence that adding courses works is the strongest we have (HIOBS).
2 · The long arc

The decline was mostly gradual, then COVID knocked enrollment down to a lower plateau

Enrollment fell steadily from 1995 to 2010, held at about 3,600–4,300 from 2012 to 2019, then dropped to roughly 2,100. Most of the earlier fall happened while nine new direct competitors launched (1995–2009) and search and social media changed how families research programs (see 2b). Removing Veterans (the teal line) lowers each year but doesn't change the shape of the decline.

OE total (leadership series)Family-paid estimate (Veterans removed)COVID break (2020)○ 2026 provisionalMarkers below the axis show changes in how families find programs
Show the data

Only years with a published value are plotted, and lines connect those years. One sheet in Historical numbers OE.xlsx shows 3,884 for 2019, not 4,259. That conflict has not been resolved. I1

Since 1995 the market grew and our price rose less than inflation, yet enrollment fell

Index, 1995 = 100. If price were the cause, the gold line would sit above inflation. It sits below it.

Show the data
Measure19952026Source
Affordable teen households100~300Internal model, Census-based I1
CPI-U (inflation)100~218BLS E8
OE Classic list price (~$2,750 → ~$4,617)100~168Internal I1
OE students (9,020 → ~2,114)100~23Internal I1

How big the share problem is

2026 OE students compared with what we would enroll at our 1995 share of affordable teen households.

This is the gap the report calls a share-of-choice collapse. Nobody should plan to recover it; it shows the scale of the problem. It also explains why trimming price wouldn't fix it: the families exist, but we aren't on their shortlist.

2b · When the competition arrived

Most of our direct competitors already existed in 1995. The last one opened in 2009.

We tracked when each of the 27 direct competitors in our Program Finder was founded. These are the providers whose summer and gap programs families compare against ours today. Founding dates show when the competitive field formed. They don't show how big each provider was, or when families started to notice it.I4E20

17 of 26core competitors were already operating when our enrollment series starts in 1995 (founding year known for 26 of 27). External
12were founded in one wave, 1983–1996. That wave built teen adventure travel and international teen service as categories. External
9launched from 1995 to 2009, when OE fell fastest (9,020 → 4,600). Both
0new core competitors since 2009. OE then held at 3,600–4,300 until COVID. Both

Founding year of each core competitor, compared with how many were operating and with OE enrollment

Hover over a row for the founder and source notes. Rows marked with * have a lower-confidence date. The lower two strips use the same years as the timeline.

Wilderness and expeditionTeen adventure travelInternational travel, service and gapOcean and sailingOutward Bound USAShaded band: the 1983–96 founding wave
Show the data

What the founding dates tell us

  • The field was mostly built before our decline began. Three of today's competitors were already running when Outward Bound came to the U.S. in 1962, and NOLS and AMIGOS followed in 1965. By 1995, 17 of the 26 were operating.
  • New entrants line up with our steepest fall. Nine competitors launched between 1995 and 2009, while OE dropped by about half. Five of the nine were international, gap-year or ocean programs, formats where Outward Bound offers little.
  • Enrollment stopped falling when new entries stopped. No core competitor has launched since 2009, and OE held at 3,600–4,300 from 2012 to 2019. The drop after 2019 lines up with COVID, the NWOBS closure and our own course cuts, not with new competitors.
  • Some of the field grew out of Outward Bound. NOLS founder Paul Petzoldt was chief instructor at Colorado Outward Bound, and Kroka's founders were Outward Bound instructors. We helped train the competition.

How to describe competitive expansion accurately

Say thisNot this
Of the 26 direct competitors we track, 17 were operating by 1995 and 9 more launched by 2009, the years OE fell fastest.“Competitors tripled since 1995.” Founding dates support about 1.5×.
Since 2009 the pressure has come from existing competitors growing, families comparing everyone side by side online, and other options for the same time and money.“New competitors keep entering the market.”
The founding wave created new categories (teen adventure travel, international service, gap year) that Outward Bound didn't fully compete in.“Competitors copied our product and took our students.” We can't trace individual families.

Founding year isn't the same as size or visibility. Many of these providers were small in 1995 and grew after search, TeenLife-style directories (late 2000s) and the Gap Year Association (2012) made them easy to find and compare. We don't have competitors' enrollment histories.

3 · The market right now

Outside data and our funnel show the same pattern: families are waiting longer to commit

The six panels below come from separate sources: three outside External and three from our own records Internal. None of these sources was built to explain OE, yet all six point the same way. Families are deciding later, asking for more help and keeping their options open.

Internal

Applications arrive closer to the course date

Median days, Classic application → start

External

Travelers book closer to departure

% of 239 travel advisors naming 1–3 months as the most common booking windowE9

Internal

Families ask for a conversation first

Consultations booked, launch window vs prior year

External

Consumer confidence keeps sliding

Conference Board Consumer Confidence Index, annual averageE4

Internal

Early summer applications remain soft

Summer applications, Oct 1–Nov 5 window, Veterans excluded

External

Household savings are thinner

U.S. personal saving rate, % of disposable incomeE6

What parents are telling outside surveys73% say inflation changed their children's summer plans. Of those, 57% scaled back a family trip and 50% scaled back camps or activities.E3 Michigan sentiment recovered to 51.0 in November 2026 but is still near its record low, and 57% of consumers say high prices have lowered their standard of living.E5 43.3% of consumers plan a vacation in the next six months, and that share is falling.E4

Market swings moved fill far more than our price changes did

Network fill rate and average price per day, 2023–2026 (same course set, NWOBS included through 2024). Fill swung 18 points in a single year. Price per day rose 8% over four years.I2

The price-elasticity study found that a 10% price increase changed fill by about 0 points (plausible range −8 to +7), and the cheapest courses did not fill better. Year-to-year market conditions mattered far more than price. That supports managing commitment friction rather than cutting price.

4 · What we control

Fewer courses, shorter courses and avoidable cancellations made the decline worse

No outside data explains this part. It comes from decisions about which courses we list, which ones run, and how long they are. These decisions can be reversed, and they're where the 2027 plan has the most leverage.

The course catalog shrank by 42%, starting with Adult courses

Listed OE courses, continuing Schools. Applications fell from 4,962 (2019) to 2,712 (2026) as listings fell.

Teen and young adultAdult

Families got fewer days for a higher daily price

Change from 2018 to 2026. The dashed line marks inflation (CPI, 2018–2025, about +28%).

Show the data
Measure20182026Change
Students4,6052,114−54%
Student program days78,01328,367−64%
Days per student16.913.4−21%
Average total cost$3,951$5,052+28%
Average price per day$235$381+62%

Source: pricing pivot (excludes Veterans and GT), which has hidden filters. Treat price per day as descriptive, not a like-for-like price index.I1

Long programs lost the most, when families are least willing to commit to them

Students, peak year vs 2025

Peak year2025

Cancellation rates rise and fall without a trend

Share of listed courses cancelled, by period

In 2026, 9 of 49 cancellations were wildfire-related, and those 9 courses had 97 applications. Teen courses were cancelled at 15.5% and filled to 88.8%. Courses outside the teen portfolio were cancelled at 31.9% and filled to 81.6%.I1

5 · Discovery and measurement

Families research differently now, and we lose track of many after they show interest

Families used to find programs through brand reputation, school fairs and referrals. Now they search, scroll social media and increasingly ask AI assistants, often without ever clicking through to our site. The outside data shows these channels changing. Our data shows that we can't follow most families from first interest to enrollment.

External

AI summaries cut clicks in half

Share of Google visits with a click on a search resultE13

External

Gap-year fairs have largely disappeared

Number of USA Gap Year FairsI1E16

External

The direct-competitor field was complete by 2009

Core competitors operating, by founding year (26 with known dates)E20

Internal

Applicants we lose track of

138

people started a 2027 application and stopped. Only 1 has a lead record, so nobody follows up. Their chosen courses carry about $750,000 in tuition. Recovering a fifth of them would be worth about $123,000, with no added media spend.I3

Why this matters for spending decisions

U.S. digital advertising is near $300B and growing 14% a year, so attention costs more each year.E14 Google now tells advertisers to rely on first-party conversion data and controlled experiments, not platform-reported attribution.I1 Until leads, consultations, applications, enrollments and aid are linked in Salesforce, we can't tell which spending produces enrolled students. Every channel debate then turns on lead volume, a measure that doesn't track enrollment.

6 · 2027–2028 outlook

No help from outside: the teen population is flat, the gap-year pool is shrinking, and families remain cautious

The core teen pool for 2027–28 is stable but about 8% below its peak

U.S. births by year, labeled with the year that group turns 15. Gold bars are the 15-year-olds of 2027 and 2028.E11

What to expect from outside conditions

Force2027–28 outlook
Teen population (13–16)Flat. Births were steady at ~3.95M a year for 2010–2016. No growth from population.
Gap-year pool (17–19)Shrinking. High school graduates peaked in 2025 and decline through 2041. Fairs and directory-driven discovery are weaker.E10
Confidence and savingsWeak, with month-to-month bounces (October CCI +7.5 points; November Michigan +6%). Plan for 2027 to resemble 2026.
Booking windowLikely to stay short. Expect demand to arrive late and plan staffing and cancellation decisions around it.
DiscoverySearch clicks keep moving into AI answers. Course pages need specific, structured information about cost, readiness, travel and outcomes so AI tools can find and cite them.
An openingParents are pushing to get phones out of kids' lives: 36 states ban or limit phones in schools (Aug 2026).E18 Phone-free, real-world challenge is what we offer, but we haven't tested that message yet.
7 · The math of stability

Holding enrollment depends mostly on things we control

OE enrollment comes down to three numbers: courses listed × share that run × students per running course. In 2026 that was 243 listed × 79.8% running × about 10.9 students (12.5 seats at 87.2% fill) ≈ 2,114 students. Use the presets or sliders to see how much a weak market year costs, and how much the controllable levers can recover.

2026: 243. 2019: 376. Add only in proven segments (teen 13–14 and 14–16, HIOBS-type course families).
2026: 20.2%. 2017–19 average: 15.6%. Teen portfolio in 2026: 15.5%.
2026: 87.2%. HIOBS: about 93%.
2024 was about −10 points across the network. 0 means a year like 2026.
Projected students
vs 2026 (≈2,114)
Courses that run
Students per running course

Illustrative. The 12.5 seats per course is derived from the 2026 totals (2,114 ÷ 194 running ÷ 87.2% fill). It isn't a planning input. Courses added outside proven segments may fill lower than the network rate. Scenario values are for discussion, not forecasts.

8 · The road

Six actions, each tied to a confirmed cause

Each action addresses a cause that both outside data and our own records confirm, and each has measures we can track by mid-season. None depends on a broad price cut or a generic awareness campaign.

1 · Capacity and mix

Add courses where they have worked

Add departures in HIOBS-type teen and young-adult course families and in network 13–14 and 14–16 products. Replace Adult routes that are repeatedly cancelled rather than adding more of them.

Track
Courses listed and run, fill, booking pace, contribution, effect on nearby departures
2 · Capacity and mix

Put cancellations under control

Use standard reason codes and record enrolled students at the moment a course is cancelled. Report separately: never offered, cancelled (by reason), ran with empty seats, and ran below full but above break-even.

Track
Cancellation rate by reason, days before start, students transferred or lost
3 · Commitment friction

Make committing feel safe

Show tuition, travel, gear and exclusions together, with aid information before families drop off. Test staged deposits and transferability. Measure net revenue before rolling anything out widely.

Track
Application completion, deposit conversion, transfers, net tuition, refunds
4 · Commitment friction

Work to the new booking window

Pace courses by days until start, not by calendar month. Trigger consultation, aid and travel support at set points. Keep viable courses open long enough for families who book late.

Track
Median days from application to start, pace vs comparable departures, consultation → enrollment
5 · Discovery

Measure enrolled students, not leads

Link lead, consultation, application, enrollment, revenue and aid records. Make every abandoned application a task someone follows up on. Use geographic and incrementality tests to see which channels produce enrollments.

Track
Record match rate, cost per enrolled student, abandoned applications recovered, measured lift
6 · Share of choice

Give families evidence for the decision

Build course pages that cover outcomes, readiness, credentials, travel handoff, payment and aid options, and what happens if plans are disrupted. Structure them so search and AI tools can find and cite them. Reposition long programs as guided pathways.

Track
Visits from organic search and AI tools, engagement on course pages, conversion by content seen
Now → Jan 2027Before booking season
  • Start cancellation reason codes and enrollment snapshots
  • Total-cost and aid view on course pages
  • Abandoned applications go to someone's follow-up queue
Feb → May 2027Main decision window
  • Pace by days until start
  • Run staged-deposit test
  • Measure response to Early Pricing
  • Hold viable courses open longer
Jun → Aug 2027Operating season
  • Cancellations reported by reason
  • First full record match rates
  • Capture consultation outcomes
Sep → Nov 20272028 inventory build
  • Expand in segments where 2027 confirmed demand
  • Retire routes that keep cancelling
  • Reposition long programs
2028Back to growth
  • Restored capacity on proven demand
  • Spending decisions based on enrolled students
  • Decisions from our own test results

Quarterly scorecard for EDs and Boards

These numbers show whether we're stabilizing. Most move before final enrollment does.

Measure2026 baselineDirection for 2027Answers
Courses listed / courses run243 / 194Up, in proven segmentsDo we have enough to sell?
Cancellation rate, by reason20.2%Toward ~15%; avoidable cancellations at zeroAre we losing courses we could have run?
Fill on courses that run87.2%Hold at 85% or above while adding coursesDid additions find demand?
Median days, application → start98Stable or longerIs commitment friction easing?
Consultation → enrollment rateNot linked yetMeasured, then risingIs help-seeking converting?
Abandoned applications recovered0 of 13820% or moreAre we keeping demand we already have?
Lead → enrollment match rateVery lowMost records linkedCan we judge marketing spend?
Aid requests / average award16.4% / $2,287Track; offer aid earlierHow hard is affordability pressing?
9 · What we're asking of EDs and Boards

Five decisions that make the plan possible

  1. Treat enrollment as a capacity and conversion problem, not only a marketing problem. Judge 2027 by courses run, cancellations avoided and conversion, alongside lead volume.
  2. Approve selective course additions for 2027–28 in proven teen and young-adult segments. Ask each School for a short plan to replace Adult routes that keep cancelling.
  3. Adopt one network standard for cancellations: reason codes and an enrollment snapshot at the moment of cancellation. Without it we can't separate weak demand from operational loss.
  4. Back controlled tests rather than broad discounts: staged deposits, transferability and total-cost presentation, measured against net revenue.
  5. Fund the connected data and the people who use it: Salesforce linking from lead to enrollment, and consultation and follow-up staff who work the late booking window.
What the evidence does not prove
  • Overlapping timelines don't prove cause. OE's decline coincides with growth in competitors, changes in discovery channels and economic cycles. We can't prove any single cause or count families lost to a particular competitor.
  • No single official 1995–2026 enrollment series exists yet. The leadership, Salesforce and pricing-pivot totals differ in scope. 2009–2011 Veterans can't yet be removed year by year. 2020 is a pandemic break.
  • 2026 is provisional. The 2,114 figure comes from a pivot with hidden filters, and the Veterans exclusion isn't final.
  • Cancellation reasons before 2018 are unreliable, and final enrollment doesn't record how many students were committed when a course was cancelled.
  • Economic indicators describe the broad market, not OE's affluent family segment specifically. Some 2026 figures are year-to-date or preliminary (for example, the November Michigan reading). Replace them with archived official figures before board publication.
  • The births-by-year chart is a simple population measure. It isn't adjusted for income, region or migration.
Sources

Internal

  1. I1 · OE Enrollment Diagnosis and 2027 Growth Strategy (2027 Planning, Nov 2026). Draws on Historical numbers OE.xlsx (leadership series), the OE historical analysis, the inventory, pricing and capacity synthesis, the 2026 operating pivot, and the 2026–27 funnel analysis. Several outside figures (travel advisors, gap-year fairs, digital ad spending, wildfire smoke) are as cited there.
  2. I2 · OE Price Elasticity: Findings and 2027 Pricing Test Plan (Nov 13, 2026). 337 same-course pairs, 2023–2026.
  3. I3 · OE Pacing Calculator v4 Supplement (Nov 17, 2026). Abandoned 2027 applications.
  4. I4 · How Did Choices for U.S. Families Explode While OBUSA Open Enrollment Shrank? (competitive landscape, Oct 2026), and the Program Finder master database (27 tracked competitors, Nov 2026). Founding dates for NOLS, Wilderness Adventures, Bold Earth, Rustic Pathways, ARCC, Overland, Road Less Traveled, Adventure Treks, Dragons, Kroka and Carpe Diem are taken from the landscape report.
  5. I5 · OBUSA Addressable Market, 1995 vs. 2026 (Oct 2026). Affordable teen households, share of choice.

External

  1. E1 · Outdoor Industry Association, 2026 Outdoor Participation Trends Report (2025 data).
  2. E2 · Afterschool Alliance, America After 3PM: The Summer Struggle for Everyday Families (May 2026); Hechinger Report on enrollment by income.
  3. E3 · National Parents Union, Inflation is Reshaping Summer (June 2026; non-probability sample, n=1,527).
  4. E4 · The Conference Board, Consumer Confidence Index, October 2026 (91.7; vacation intentions 43.3%), via AP coverage; annual averages from I1.
  5. E5 · University of Michigan Surveys of Consumers. November 2026 preliminary 51.0, via ABA Banking Journal. May 2026 low of 44.8 from I1.
  6. E6 · U.S. Bureau of Economic Analysis, Personal Saving Rate (2026 is year-to-date).
  7. E7 · Federal Reserve, Survey of Household Economics and Decisionmaking (2025 data).
  8. E8 · U.S. Bureau of Labor Statistics, Consumer Price Index (CPI-U).
  9. E9 · 2026 travel advisor booking-window survey (n=239), as cited in I1.
  10. E10 · WICHE, Knocking at the College Door, 11th ed. (Dec 2024).
  11. E11 · CDC National Center for Health Statistics, National Vital Statistics System, births (final counts 2007–2016).
  12. E12 · Federal Reserve Bank of Minneapolis, Have U.S. consumers gone K-shaped? (2026), comparing Moody's and BLS estimates.
  13. E13 · Pew Research Center, Google users are less likely to click on links when an AI summary appears (July 2025).
  14. E14 · IAB/PwC Internet Advertising Revenue Report, 2025 (~$300B, +13.9%), as cited in I1.
  15. E15 · Apple, App Tracking Transparency.
  16. E16 · Gap Year Association provider directory; IECA, Seven Gap Year Trends to Know for 2026.
  17. E17 · NOLS gap-year program marketing (as reviewed in I1).
  18. E18 · Ballotpedia, K-12 cellphone policy tracking (state count as of Aug 2026).
  19. E19 · BEA, Outdoor Recreation Satellite Account (2024: $696.7B, 2.4% of GDP).
  20. E20 · Founding dates from provider sites and directories: ActionQuest (1986; Seamester 1998); Broadreach (1993); Global Works (1989); VISIONS (1988); Apogee (2001); Travel For Teens (2003); NCOAE (2009); Chewonki (1915); AMIGOS (1965); Putney (1951); GLA (2004); Moondance (1996). Lower confidence: Cottonwood Gulch (1926, directory listing only); EF Ultimate Break (~2003 as EF College Break, from a single review). Not found: Breakwater Expeditions.