WHAT ACTUALLY HAPPENS AFTER STUDENTS ENROLL: DATA DICTIONARY AND CAVEATS Opportunity Data, Benazir Rowe, PhD. Page version 1.1 and methodology note version 1.3, November 19, 2026 (first released November 17, 2026): https://opportunitydata.org/briefs/what-actually-happens-methodology.pdf READ FIRST 1. Every "of_100" figure is an estimate built by multiplying rates from different federal sources and cohorts. No single group of students was followed from enrollment to earnings. Completion, transfer, departure: IPEDS Outcome Measures, all students entering in 2017-18, status 8 years after entry. Transfer outcomes: NSLDS (federal aid recipients entering 2015-16, measured 2023-24), applied to the Outcome Measures transfer share. NSLDS records completion at the starting college less often than Outcome Measures (19 vs 32 per 100 at community and technical colleges nationally); students whose first credential went unrecorded appear as transfers, so the share of transfers who finish may be overstated. Work and earnings: College Scorecard field-of-study file (Treasury), federal aid recipients completing in 2020-21 or 2021-22, earnings in calendar 2022-23, 2024 dollars. Completers still enrolled are excluded. Eight-year entrant measures: Scorecard institution file, aid recipients entering 2013-14 or 2014-15, not enrolled. ACS measures: 2020-24 five-year microdata, people not enrolled in school. 2. The earnings stage covers completers at the starting college only. Students who finish after transferring or who leave are not in it. 3. Two benchmarks. "Out-earn a high school graduate" (the of_100 ... above_hs columns, *_working_above_hs_share, and the eight-year share) uses Treasury counts that match ONE NATIONAL figure (tested against published counts; the Scorecard documentation describes a single comparison figure): the median earnings of high school graduates aged 25 to 34 ($35,678, Scorecard EARN_THR_NAT). The premium columns use each STATE's median (EARN_THR_STATE). Both medians are ACS 2020-24 five-year estimates. The published values fit medians for high school graduates with positive earnings (our test; the Scorecard documentation describes workers in the labor force). The comparison group is older, so first-year premiums typically understate later ones. In states where the high school median is below the national value, the out-earn shares are lower than a state comparison would give; where it is above, higher. 4. "No earnings" means no earnings recorded in federal tax records (Scorecard) or none reported in the past 12 months (ACS). It is not the unemployment rate. 5. Jobs below the credential uses one BLS typical entry education per occupation (Employment Projections 2024-34). It does not measure skills used. The ACS does not identify certificates. 6. Figures describe students and colleges; they do not estimate what a credential causes anyone to earn. Differences reflect who enrolls as well as what colleges do. 7. Sectors: public community and technical colleges = public institutions predominantly awarding certificates or associate's degrees (plus unclassified public two-year and less-than-two-year campuses); public universities = other public institutions; private nonprofit and for-profit include all levels. Fully online institutions are a separate national group. Puerto Rico and territories are excluded. Sector follows the federal classification of the award a college mainly confers (PREDDEG), so Purdue University Global, Utah Valley University, and Weber State University are counted with community and technical colleges. Colleges that do not report Outcome Measures (including Tennessee's colleges of applied technology and many certificate-only schools) are left out of every stage, including earnings (since version 1.2). Sector codes in the drill-down and supporting files: all = all colleges, pub2 = public community and technical colleges, pub4 = public universities, pnp = private nonprofit, fp = for-profit, online = fully online institutions. level_code: 1 = certificate, 2 = associate's degree, 3 = bachelor's degree. 8. shown_on_page = FALSE when fewer than 300 students entered or reported credentials cover under 80 percent of awards. Those rows are less reliable; use with care. 9. Scorecard earnings pooled at the parent institution (six-digit OPEID) repeat on branch campuses; each record is counted once (field of study: campus with most completers; institution: main campus or, failing that, the campus with the largest entering cohort). A pooled record that spans campuses in more than one state or sector is counted wholly where it is kept, so a few state and sector figures include branch earnings from elsewhere. 10. Counts and dollar values are written with a decimal point (for example 11300.0); they are whole numbers. 11. completers_ipeds counts awards in one IPEDS award year; Scorecard earnings pool two completion years. A row can therefore show completers_ipeds = 0 alongside published earnings. RIGHTS AND REUSE (c) 2026 Opportunity Data. The text, charts, and derived estimates on this page, in the methodology note, and in the downloadable files are the work of Opportunity Data. You may quote the text, reproduce the charts unchanged, and use the files for analysis, including your own charts, for news reporting, research, teaching, and public policy work, with the credit "Source: Opportunity Data, opportunitydata.org." Reselling the files, republishing them as a dataset, or building them into a commercial product requires written permission: info@opportunitydata.org. Values taken directly from federal sources (IPEDS, the College Scorecard including its NSLDS and Treasury measures, the American Community Survey, BLS Employment Projections) are public and remain free to use from their original publishers. Cite as: Rowe, Benazir. What Actually Happens After Students Enroll. Opportunity Data, November 17, 2026 (version 1.1, November 19, 2026). https://opportunitydata.org/what-actually-happens.html COLUMNS state State name, or United States. state_code Postal code; US for the national row. sector Sector (see READ FIRST 7). shown_on_page Whether the row met the coverage rules for the chart (READ FIRST 8). entrants_ay2017_18 Degree- and certificate-seeking undergraduates entering in 2017-18 in the Outcome Measures adjusted cohorts, all four entry groups. institutions Institutions reporting Outcome Measures in the cell. of_100_earned_credential_at_starting_college_8yr Per 100 entrants: received an award from the starting college within 8 years (Outcome Measures). of_100_transferred Per 100 entrants: no award at the starting college, later enrolled at another college (Outcome Measures). of_100_still_enrolled Per 100 entrants: no award, still enrolled at the starting college at year 8. of_100_left_no_record_elsewhere Per 100 entrants: no award, subsequent enrollment unknown to the starting college ("left, no record elsewhere"). of_100_transferred_then_earned_credential Per 100 entrants: estimated transfers who earned a credential at a later college within 8 years (NSLDS split; may be overstated, READ FIRST 1). of_100_transferred_then_left Per 100 entrants: estimated transfers who withdrew with no later enrollment record, plus status unknown (NSLDS split). of_100_transferred_still_enrolled Per 100 entrants: estimated transfers still enrolled at a later college at year 8 (NSLDS split). of_100_credential_anywhere_8yr Per 100 entrants: credential at the starting college plus estimated credential after transferring. of_100_credential_and_earn_above_hs_1yr Per 100 entrants: estimated completers at the starting college who, one year after completing, worked and earned more than the national high school benchmark (READ FIRST 3). of_100_credential_and_earn_at_or_below_hs_1yr Per 100 entrants: estimated completers who worked and earned at or below the national high school benchmark. of_100_credential_and_no_earnings_1yr Per 100 entrants: estimated completers with no recorded earnings and not enrolled. of_100_credential_earnings_not_reported Per 100 entrants: completers in credential levels whose programs are too small for Scorecard to publish counts. entrants_8yr_after_entry_share_working_above_hs Of aid recipients entering 2013-15 and not enrolled 8 years after entry, the share who worked and earned above the national high school benchmark: (1 - share with no earnings) x GT_THRESHOLD_P8. GT_THRESHOLD_P8 itself is a share of those working and not enrolled; this column rescales it to everyone not enrolled. entrants_8yr_after_entry_share_no_earnings Share of the same group with no recorded earnings. average_annual_tuition_fees Published in-state tuition and fees per year, in dollars, 2025-26 (TUITIONFEE_IN; program-year reporters excluded), weighted by entering cohort. average_annual_net_price Average net price per year, in dollars (cost of attendance including living costs, minus grants), for first-time full-time aid recipients, 2024-25, weighted by entering cohort. state_hs_median_earnings_25_34 State median earnings of high school graduates aged 25 to 34 (national for the US row), Scorecard EARN_THR_STATE / EARN_THR_NAT; used for the premium columns only. cert_share_of_completers Share of the sector's completers earning certificates (awards reported to IPEDS for one award year, IPEDSCOUNT2 in the Scorecard file). cert_reported TRUE when at least 100 completers are counted for work and 50 for the benchmark. When FALSE, any earnings values in the row are shown for reference only; the of_100 figures count that level as earnings not reported. cert_completers_no_earnings_share Of completers not enrolled, share with no recorded earnings one year after completing. cert_working_above_hs_share Of working completers, share earning more than the national high school benchmark (READ FIRST 3). cert_median_earnings_1yr Median of program medians, weighted by working completers, one year after completing, 2024 dollars. cert_premium_over_hs_1yr Median earnings minus the state median for high school graduates aged 25 to 34 (EARN_THR_STATE). First-year, graduates who work only; not a causal estimate. cert_tuition_fees_nominal_length Annual tuition and fees times nominal length (certificate 1 year, associate 2, bachelor 4). Understates cost for students who take longer; ignores grants. cert_total_net_price_nominal_length Annual net price times nominal length. Includes living costs. cert_of_100_starters_reach_above_hs Per 100 entrants in the sector: estimated number who earn certificates at the starting college and earn above the national high school benchmark a year later. assoc_share_of_completers Share of the sector's completers earning associate's degrees (awards reported to IPEDS for one award year, IPEDSCOUNT2 in the Scorecard file). assoc_reported TRUE when at least 100 completers are counted for work and 50 for the benchmark. When FALSE, any earnings values in the row are shown for reference only; the of_100 figures count that level as earnings not reported. assoc_completers_no_earnings_share Of completers not enrolled, share with no recorded earnings one year after completing. assoc_working_above_hs_share Of working completers, share earning more than the national high school benchmark (READ FIRST 3). assoc_median_earnings_1yr Median of program medians, weighted by working completers, one year after completing, 2024 dollars. assoc_premium_over_hs_1yr Median earnings minus the state median for high school graduates aged 25 to 34 (EARN_THR_STATE). First-year, graduates who work only; not a causal estimate. assoc_tuition_fees_nominal_length Annual tuition and fees times nominal length (certificate 1 year, associate 2, bachelor 4). Understates cost for students who take longer; ignores grants. assoc_total_net_price_nominal_length Annual net price times nominal length. Includes living costs. assoc_of_100_starters_reach_above_hs Per 100 entrants in the sector: estimated number who earn associate's degrees at the starting college and earn above the national high school benchmark a year later. bach_share_of_completers Share of the sector's completers earning bachelor's degrees (awards reported to IPEDS for one award year, IPEDSCOUNT2 in the Scorecard file). bach_reported TRUE when at least 100 completers are counted for work and 50 for the benchmark. When FALSE, any earnings values in the row are shown for reference only; the of_100 figures count that level as earnings not reported. bach_completers_no_earnings_share Of completers not enrolled, share with no recorded earnings one year after completing. bach_working_above_hs_share Of working completers, share earning more than the national high school benchmark (READ FIRST 3). bach_median_earnings_1yr Median of program medians, weighted by working completers, one year after completing, 2024 dollars. bach_premium_over_hs_1yr Median earnings minus the state median for high school graduates aged 25 to 34 (EARN_THR_STATE). First-year, graduates who work only; not a causal estimate. bach_tuition_fees_nominal_length Annual tuition and fees times nominal length (certificate 1 year, associate 2, bachelor 4). Understates cost for students who take longer; ignores grants. bach_total_net_price_nominal_length Annual net price times nominal length. Includes living costs. bach_of_100_starters_reach_above_hs Per 100 entrants in the sector: estimated number who earn bachelor's degrees at the starting college and earn above the national high school benchmark a year later. DRILL-DOWN FILES what-actually-happens-fields-by-state-and-sector.csv (one row per state x sector x credential level x 4-digit CIP field) Earnings stage only: Outcome Measures do not report completion by field, so these rows describe completers, not everyone who started. Built from the same Scorecard field-of-study records as the main table (pooled records counted once). p_not_working is shown when at least 30 completers are counted for work; p_above_hs_if_working (share above the national high school benchmark), median_earnings, and premium when at least 30 working completers are counted against the benchmark. median_earnings is the median of program medians weighted by working completers, one year after completing, 2024 dollars. premium = median_earnings minus the state benchmark (national benchmark in the United States rows). colleges = campuses with a program in the field. Columns: STABBR (state, US for national), sector and level_code (READ FIRST 7), cip4 (4-digit CIP field code), field (CIP title), completers_ipeds (awards in one IPEDS year), work_base (completers counted for work), threshold_base (working completers counted against the benchmark). Rows with work_base under 100 are tagged "few completers" on the page. Not causal estimates. what-actually-happens-colleges.csv (one row per campus with at least 50 Outcome Measures entrants) Each college's own figures; no stages are multiplied at this level. entrants: degree- and certificate-seeking undergraduates who entered the college in 2017-18 (Outcome Measures adjusted cohort), all four Outcome Measures groups (first-time and transfer-in students, each full-time and part-time). p_award, p_transfer, p_still, p_unknown: Outcome Measures, 2017-18 entrants, eight years. tr_p_completed: of the college's federal aid recipients who entered in 2015-16 and transferred before earning a credential there, the share who completed at a later college within eight years (NSLDS); may be overstated (READ FIRST 1). Pooled across campuses of a multi-campus institution and repeated on each one. tuition_year: in-state tuition and fees per year, in dollars, 2025-26; net_price_year: average net price per year, in dollars, 2024-25. cert_/assoc_/bach_ columns: the college's completers one year after completing, same thresholds as the field file (*_completers_ipeds: awards in one IPEDS year; *_p_not_working: share with no earnings; *_p_above_hs_if_working: share of workers above the national high school benchmark). earnings_pooled_elsewhere = TRUE when the campus's Scorecard earnings are pooled under another campus of the same institution (six-digit OPEID); its earnings columns are then empty. College figures rest on small counts more often than state figures and move more from year to year. Some campuses share a name within a state; UNITID tells them apart. SUPPORTING FILES what-actually-happens-completion-by-entry-group.csv (one row per sector x Outcome Measures entry group, national) entry_group: FTFT = first-time full-time, PTFT = first-time part-time, FTNFT = transfer-in full-time, PTNFT = transfer-in part-time. entrants: adjusted cohort, 2017-18. p_award, p_transfer, p_still, p_unknown: shares of entrants with an award, transferred, still enrolled, or status unknown 8 years after entry. what-actually-happens-acs-state-measures-2020-24.csv (one row per state x education group, ACS 2020-24 PUMS) group: hs = high school diploma or equivalent, some_college = some college and no degree, associate, bachelor. People not enrolled in school. unemp_22_27: unemployed share of the civilian labor force, ages 22 to 27. p_no_earnings_25_34: share with no earnings in the past 12 months, ages 25 to 34. p_above_hs_if_earning_25_34: of those with earnings, the share above the state median for high school graduates. underemp_hs_or_less_22_27 (associate and bachelor only): share of employed people in occupations whose BLS typical entry education is a high school diploma or less (the page measure). underemp_22_27: the broader measure, occupations whose typical entry education is below the worker's degree. _se: standard error from the 80 replicate weights. _n: unweighted records in the denominator (whole numbers). Rates resting on fewer than 100 records are not shown on the page. what-actually-happens-sensitivity-branch-duplication.csv (one row per sector, national) j_award_above_no_dedup: per 100 entrants, finish and out-earn the national high school benchmark when every branch copy of a pooled Scorecard record is counted. p8_above_hs_of_all_no_dedup: the eight-year share, same rule. Compare with the deduplicated values in the state table.