Back to Articles|Published on November 7, 2026|29 min read

Docket Article

IPEDS OPE ID Crosswalk: One School or Five Campuses?

A 2026 data report on the IPEDS OPE ID crosswalk for vocational schools: UnitID vs 6- and 8-digit OPE ID, parent/child finance fields, and a campus-to-operator worksheet.

By Docket Research

IPEDS OPE ID Crosswalk: One School or Five Campuses?

Summary

  1. 01UnitIDs identify IPEDS reporting units, eight-digit OPE IDs identify locations, six-digit roots identify Title IV institutions, and legal operators identify the companies a buyer would approach.
  2. 02A shared six-digit OPE ID root establishes a shared PPA holder. Different roots can still belong to the same owner, so ownership requires registry or filing evidence.
  3. 03Scorecard outcomes can repeat across UnitIDs sharing a root, while finance values can combine parent and child institutions. Check each field's reporting level before summing or comparing.
  4. 04IPEDS does not capture every state-licensed vocational school. Non-Title IV sites may appear only in state registries, and additional locations may be included inside another UnitID.
  5. 05Record the collection year and crosswalk vintage, grade ownership claims, and reopen affected rows after structural events. The article does not estimate the number of independent operators.
Inside this article
  1. 01Executive Summary
  2. 02Introduction and Background
  3. 03Defining the Mandate and the Education Population
  4. 04UnitID, OPE ID and the One-to-Many Problem
  5. 05Choosing the Denominator: Institution, Campus and Operator
  6. 06Parent and Child Finance Reporting
  7. 07Reconciling to the Legal Operator
  8. 08A Fictional Crosswalk Worksheet (Hypothetical Example)
  9. 09Data Analysis and Evidence
  10. 10Implications and Future Directions
  11. 11Frequently Asked Questions (FAQs)
  12. 12Conclusion

Executive Summary

There is no one-to-one key between the two identifiers that analysts reach for first. The National Center for Education Statistics (NCES) states that IPEDS UnitID and OPE ID values "are not a one-to-one match" [1]. The College Scorecard crosswalk, built jointly by NCES and Federal Student Aid (FSA), links the two systems for 2000-01 through 2022-23 [2]. One American Council on Education (ACE) analysis, drawn from a College Scorecard data file updated September 14, 2022, counted 6,681 unique UnitIDs against 5,253 unique OPEIDs (a historical snapshot that is not comparable with the 2025-26 Title IV count) [3], and found 345 OPEIDs tied to more than one UnitID, roughly 70 percent of them for-profit [3].

For a private-equity origination team, "one school" can mean four different things: a UnitID (an IPEDS reporting unit), a six-digit OPE ID root (a Title IV institution), an eight-digit OPE ID (a location), or a legal operator. In the Fall 2025 provisional IPEDS collection, NCES counts 5,711 Title IV institutions, 2,098 of them for-profit and 1,327 of those at the less-than-2-year level [4]. Those are reporting units, not operators, and NCES separately excludes 498 institutions that were reported exclusively by a parent institution [5]. This report measures no operator counts; it supplies the bridge for producing one.

Finance data add a second trap. NCES warns that duplicate data submitted for each school in a system or corporate entity made totals overstated [6]. On the for-profit form, a parent can report the sum of parent and child data for some parts [7], so each field's reporting level must be read before summing. A 2025 peer-reviewed note on IPEDS administrative units examined only nonprofit institutions and says the for-profit sector "remains to be explored" [8], so its numeric results should not be carried over to vocational schools.

The legal operator is not an IPEDS field. It has to be reconciled from state licensing records, which are often location-based (Texas requires a certificate of approval for each location [9]; Ohio's rule says the board does not recognize branch campuses [10]), and from federal ownership reporting that begins at a 5 percent interest [11]. Parent relationships should stay sourced hypotheses until a registry or filing verifies them, and a merger, closure or ownership change should trigger a recheck.

5,711Title IV institutions in the 2025-26 provisional IPEDS table, excluding administrative offices
1,327For-profit less-than-2-year Title IV institutions in the 2025-26 provisional IPEDS table
345OPEIDs associated with multiple UnitIDs in ACE's September 2022 snapshot
1,187UnitIDs connected to those multi-UnitID OPEIDs in ACE's September 2022 snapshot

Introduction and Background

This report is written for U.S. private-equity origination analysts who map vocational and career school platforms and add-ons. The decision at stake is narrow but costly when wrong: whether several institution or campus records describe one provider or several separate acquisition candidates, and which reported metrics can be compared across them. IPEDS, the Integrated Postsecondary Education Data System, is organized around institutions as reporting units. Federal Student Aid organizes Title IV eligibility around the Program Participation Agreement (PPA) and its OPE ID. A company that owns schools appears in neither as a first-class record, so the owner has to be reconstructed from other evidence. Getting that reconstruction wrong in either direction has a cost: a missed affiliation hides a roll-up that already exists, while an assumed affiliation inflates a platform's apparent footprint.

A 2025 Research in Higher Education note by John Cheslock documents a related unit-of-analysis problem in IPEDS finance data, where reliance on parent-child indicators captures only a minority of administrative-unit observations [8]. That study examined nonprofit institutions only [8], which leaves room for a for-profit and vocational worksheet.

The report proceeds in the order an analyst would work: the mandate and population; UnitID, six-digit and eight-digit OPE IDs, and branch handling; the choice of denominator; parent and child finance reporting; reconciliation to a legal operator; a fictional worksheet (labeled as hypothetical); and a data section quantifying the reporting-unit universe as of November 2026. Sources were checked as of 2026-11-08, and dated sources are stated as such. The method here is general and does not depend on any vendor.

Defining the Mandate and the Education Population

Who is in the federal data, and who is not

IPEDS completion is mandatory for institutions that participate, or apply to participate, in Title IV [12]. For institutions not eligible for Title IV, participation is voluntary [13]. A vocational school that does not take federal aid is therefore largely absent from IPEDS, and its absence says nothing about its size.

State licensing is a separate gate. Texas rules define a "Title IV school" as a career school that participates in federal student financial aid, a status distinct from state approval [9]. New York states that licensed private career schools meet the requirements for state authorization for Title IV eligibility [14], which makes licensure a precondition, not evidence of participation. Federal regulation requires that a State have a process to review and act on complaints before an institution counts as legally authorized [15]. State registries can therefore show schools that IPEDS will never contain. Florida's Commission for Independent Education, for example, reported 740 licensed institutions that offer only non-degree programs [16], without stating how many take Title IV aid, and New York says about 500 private career schools are licensed there [17].

What counts as "vocational" in Title IV terms

For for-profit schools, the regulatory anchor is the proprietary institution of higher education: one that is not a public or private nonprofit institution [18] and that provides an eligible program of training to prepare students for gainful employment in a recognized occupation [18]. FSA's handbook adds that a proprietary institution is always a private, for-profit institution [19]. Cosmetology schools are named by NCES among the non-degree-granting institutions in the IPEDS universe [20]. Not every career school is proprietary, so the control field matters: in 2021-22, private for-profit institutions were the only control type conferring a majority of awards below the bachelor's level [21]. A mandate that says "vocational" should state its filters explicitly:

  • Control: for-profit only, or for-profit plus nonprofit and public career and technical providers.
  • Level: less-than-2-year, 2-year, or any institution with sub-baccalaureate awards.
  • Title IV status: participants only, or participants plus state-licensed non-participants (which requires registry data, not IPEDS).
  • Geography: states only, or also other jurisdictions, which some NCES tables include.
  • Modality: campus-based, online, or both, because modality can split one OPE ID across UnitIDs.

UnitID, OPE ID and the One-to-Many Problem

What each identifier measures

The UnitID is the IPEDS identifier for a reporting institution; the Scorecard's institution files are provided at the UNITID level [22]. FSA's handbook glossary defines the OPEID as an eight-digit number assigned upon approval to participate in Title IV [23]. The number is a six-digit root followed by a two-digit suffix that identifies branches, additional locations and other entities of the eligible institution [22]. Related institutions share a common six-digit root tied to the main campus, the entity that signs the PPA [22].

Several conventions are worth coding into a crosswalk:

  • Root and suffix: the first six digits identify the institution and the last two identify the location [24]. The source is a 2013 FSA draft notice, so confirm against current FSA material.
  • Main campus: the Scorecard documentation (October 2026 version) says main campus status is derived from the OPEID [25].
  • Overflow: the same 2013 notice says that past 99 additional locations the first digit is incremented [24], so a leading zero is not a reliable test.
  • Parent and child: one researcher noted that parent institutions typically have OPEIDs ending in "00" [26], using 2015-16 data, which is a heuristic and not a rule.

Branch campus versus additional location

The federal definitions decide whether a site can ever be its own row. Under 34 CFR 600.2, an additional location participates in Title IV only through the certification of the main campus [27], so it has no PPA of its own. A main campus is certified as such by the Department and the institution's accrediting agency [27]. A branch campus must be approved by the Secretary of Education as a branch campus [27] and must have its own budgetary and hiring authority [27]. FSA's handbook says an additional location cannot be at the same address as a main location [28], and that a proprietary or vocational branch must be designated a branch for two years before seeking main-campus or freestanding status [19]. The accreditor ACCSC uses the same branch approval language for its Title IV institutions [29].

NCES explains that one reporting option is for the main campus to report combined data for itself and any additional locations [13]. The Scorecard documentation notes that institutions may instead list additional locations as separate entities [22]. A physical site can therefore be (a) invisible inside a main campus row, (b) a suffix on a shared root OPE ID, or (c) a separate UnitID, depending on choices the school made.

Why the mapping runs in both directions

The Scorecard documentation states that there is not necessarily a one-to-one relationship between UNITID and the eight-digit OPEID [22]. A UNITID may aggregate several OPEIDs when locations combine their IPEDS reporting; in rarer cases a UNITID covers only part of an OPEID, such as a location that reports campus-based and online operations under separate UNITIDs [22]. In the ACE analysis, an institution with multiple branch campus locations may have each self-report to IPEDS, while loan-system data group them under one OPEID [3].

Reporting practice also changes. NCES gives the example of a university that shifted back and forth between campus-level and system-level reporting [1], and the Scorecard documentation warns that UNITIDs for an institution are not always consistent over time [22]. Any panel built on UnitID alone will therefore show spurious openings, closures and enrollment jumps.

Six-digit versus eight-digit: where outcomes live

The Scorecard documentation says that although some institutions report at the campus level (eight-digit OPE ID), data produced for the site are rolled up to the institution level (six-digit OPE ID) [22]. IPEDS institutions sharing a common six-digit OPEID are assigned the same sum, student-weighted average or median for loan-system or tax-derived measures [22], a limitation shared by the cohort default rate [22]. ACE illustrates the consequence: campuses sharing an OPEID show different net prices yet the same earnings value [3], and it recommends defining an institution at the OPEID level when outcomes are the object [3]. One researcher likewise advises analysts to note when data are aggregated across separate UnitIDs [26]. A campus-level earnings or repayment figure from the Scorecard is therefore generally a six-digit group figure repeated across child rows, and must not be treated as a campus measurement.

One caution applies to chain mapping. The Scorecard documentation notes that some large public and for-profit systems assign different six-digit OPEIDs to each campus [25]. A shared six-digit root can prove a shared PPA holder; different roots do not prove different owners.

The Scorecard crosswalk file

The College Scorecard data page, last updated November 2, 2026, offers crosswalk files for 2000-01 through 2022-23 that link the Department's OPEID to an IPEDS UNITID [2]. The documentation says the crosswalks were built jointly by NCES and FSA with institutional history back to 2000-01 [22], combining FSA participant-system data with IPEDS mappings [22]. FSA's handbook glossary notes that its Postsecondary Education Participants System was retired in 2024 [23], so analysts should record which crosswalk vintage they used and not assume it tracks current participation. The Scorecard also supplies MAIN, a flag where 1 is a main campus and 0 is not [22], and NUMBRANCH, a count of branch campuses [22]. Most Scorecard documentation passages cited here come from the June 2024 version of the technical documentation; an October 2026 version also exists, and passages should be rechecked against it. The current data dictionaries were not available for verification as of November 2026, so variable names should be copied from the dictionary for the year in use.

A shared six-digit root can prove a shared PPA holder; different roots do not prove different owners.

Choosing the Denominator: Institution, Campus and Operator

An origination team needs at least four denominators, and errors come from mixing them. Table 1 below sets out the bridge from campus rows to verified legal operators. It contains no counts, because none has been measured for a named mandate; each "count" cell is a slot for the analyst's own dated run.

Table 1. Denominator bridge (template; counts to be measured, none reported here)

StepUnit countedWhere it comes fromMain failure modeCount to record
1State-licensed sites in the mandateState registries, one per jurisdictionLocation-based approvals counted as institutions; non-Title IV sites mixed inn = measured, dated
2Eight-digit OPE ID rows (locations)Crosswalk and Title IV participant dataAdditional locations have no PPA of their ownn = measured, dated
3Distinct IPEDS UnitIDsIPEDS directory (IC-H)Satellites reported inside another UnitIDn = measured, dated
4Distinct six-digit OPE ID rootsDerived from the eight-digit valuesDifferent roots may share an ownern = measured, dated
5Distinct Title IV institutions after collapsing parent and childIPEDS parent/child flags plus OPE rootsChild institutions are reported in whole or part by a parentn = measured, dated
6Verified legal operatorsRegistry or filing evidence for each PPA holderBrand or shared root treated as proof of common ownershipn = measured, dated
7Candidate acquisition unitsAnalyst judgment on steps 5 and 6Platform and add-on logic applied to the wrong denominatorn = measured, dated

The order is not strictly monotonic: a registry can show more sites than IPEDS has campuses (non-Title IV sites), and one operator can hold several PPAs. The discipline is to label every count with its unit. A statement such as "40 campuses" must never be repeated later as "40 operators" or "40 targets". Program counts are a further hazard: CIP (Classification of Instructional Programs) codes identify instructional program specialties, so completions are program measures that say nothing about the number of operators.

The headline universe is itself a mix of reporter types. NCES describes its 2025-26 universe as 5,779 Title IV institutions, administrative offices and U.S. service academies [5]; the 5,711 institutions counted elsewhere in this report are the Title IV institutions within that total, excluding administrative offices. A reader who divides enrollment by "institutions" without that qualification is dividing by a number that already embeds reporting choices.

Parent and Child Finance Reporting

What the terms mean

The IPEDS glossary defines a parent institution as one that reports some or all data for another institution [30]. NCES says the Finance component lets institutions covered by a joint audit report some or all of their finance data together [6], and its illustration is a set of beauty schools owned by one entity and audited together as one company [6]. The warning is explicit: some institutions in a system or corporate entity each submitted the same data [6].

How the for-profit forms split the fields

On the 2025-26 degree-granting for-profit form, the parent reports the sum of parent and child data for Parts A, B and G [7] and parent data only in Parts C, D and E, while child institutions report their own data for those parts [7]. Under full parent/child reporting, which is allowed only where branch campuses operate under the main campus's PPA, the parent reports all finance data for all institutions combined [6].

Table 2 below translates these rules into a warning grid. It works at the level of the form part, not the variable; variable names must come from the current data dictionary.

Table 2. Finance-reporting levels on the 2025-26 for-profit form (read before summing)

Form part (as described by NCES)Reporting level in partial parent/childSafe to sum across child rows?Screener rule
Balance sheet and changes in equity (Parts A and B, with Part G on the degree-granting form)Parent-level: the parent reports the sum of parent and child dataNo. Child rows carry no separate value, and a join may repeat the parent's totalAttach once to the finance-reporting parent; never allocate to children without a stated basis
Scholarships and fellowships (Part C)Each institution reports its ownYes, if each row is confirmed as its own reportSum at the six-digit root only after checking for duplicates
Revenues and expenses (Parts D and E)Parent reports parent data only; children report their ownYes, with the parent's own value counted onceCompare against the audited group total where one exists
Full parent/child combined reportParent reports every institution combined; children report noneNo. Adding children to the parent double countsTreat the parent as the only valid row
Jointly audited institutions with separate PPAsEach files its own submissionCheck for identical values before summingFlag identical figures across rows as probable duplicates

Three corollaries follow:

  • Mixed levels: a per-campus margin built from a parent-level numerator and a child-level denominator is not meaningful.
  • Shared PPA: combined reporting requires every institution in the combined report to share the same PPA [31], which gives the shared root a second use as a plausibility test for a parent/child claim.
  • Reviewed flags: institutions cannot set up the relationship on their own and must call the IPEDS Help Desk [6], so a flag reflects a reviewed relationship.

The research record

Before comparing two schools, confirm that both rows come from the same fiscal year and the same reporting level.

The academic record is thin for the for-profit case. The Cheslock note finds that collapsing data to the Title IV level incorporates dollars from only a subset of administrative offices [8], but it examined nonprofit institutions only [8], and its numeric findings, including its 79 administrative offices affiliated with nonprofit Title IV institutions [8], should not be applied to for-profit schools. Separately, a 2017 analysis by Robert Kelchen noted that for-profit college chains typically report using parent/child relationships [26], a statement about OPEID-based loan-system groupings in 2015-16 and not about current IPEDS finance flags.

Figure 01
Finance reporting: partial versus full parent/child
Partial parent/child reporting
  • Attach parent-level balance sheet and equity totals once to the finance-reporting parent.
  • Sum scholarships and fellowships only after confirming each row reports its own data and checking for duplicates.
  • For revenues and expenses, count the parent's own value once and include children's own reports.
Full parent/child reporting
  • The parent reports all institutions combined; children report no finance data.
  • Adding children to the combined parent report double counts.
  • Treat the parent as the only valid finance row.

Before comparing two schools, confirm that both rows come from the same fiscal year and the same reporting level.

Reconciling to the Legal Operator

Identifiers describe reporting and eligibility; they do not name the company that owns the school. Three layers of evidence are available, in rising strength: the public brand, the PPA holder and shared root OPE ID, and a legal entity verified in a registry or official filing.

Federal evidence

FSA's handbook says the Department requires an authorized representative of any entity with direct or indirect ownership of a private institution to sign the PPA in appropriate cases [19]. Schools must have a current PPA to participate [19]. Eligible institutions must generally report listed changes, including ownership changes, no later than 10 days after they occur [11]; one trigger is a person or entity acquiring at least a 5 percent ownership interest [11], although changes of at least 5 percent but under 25 percent are generally reported quarterly based on the institution's fiscal year rather than within 10 days, and before a planned change in ownership is noticed to the Department, all unreported ownership changes of 5 percent or more must be reported before the 90-day notice under 34 CFR 600.20 is submitted, while an institution owned by a publicly traded corporation has a separate duty to report such changes when it notifies its accrediting agency, but no later than 10 days after it learns of the change [11]. A private for-profit institution that undergoes a change in ownership and control ceases to qualify as eligible unless the Secretary continues its participation [32]; for closely held corporations, one threshold is acquiring more than 50 percent of voting stock [32], and the regulation defines the "parent" as the legal entity that controls the institution directly or indirectly [32].

Two cautions keep these facts in proportion. The glossary description of the Eligibility and Certification Approval Report lists accreditor, state authorization, staff, additional locations and eligible vocational programs [23]; ownership is not among the items listed, so the report should not be assumed to settle ownership. And FSA said in 2024 that it would post participation statuses and PPAs on its data center website [33], but those pages were not available for verification as of November 2026, so file contents and update dates remain unverified. When one school buys another and both stay open, FSA's handbook says they keep separate OPEID numbers unless they apply to merge [28], so a change of owner does not merge identifiers.

State and accreditor evidence

State records often name the operator more directly than federal files, but their units differ by jurisdiction. The examples are illustrative, not a 50-state survey:

  • Texas: a school must obtain a certificate of approval for each location [9], the certificate is issued to the school owner and is not transferable to a new owner [34], rules recognize an identified owner, director or owner designee as a school authorized official [9], and notice of a change of ownership is due at least 45 days beforehand [9].
  • Florida: each institution must be a Florida corporation or limited liability partnership, or registered as a foreign corporation [35], the rules define additional locations as locations other than auxiliary classroom space that support a licensed institution [35], and they define change of ownership as a transfer of issued or outstanding stock or similar interests [35].
  • Ohio: the board maintains a database of registered career colleges and schools [36], and its certificates of registration are not transferable [10].

Accreditors provide another cross-check. ACCSC defines affiliated schools as a main school and its branch campuses, or a group of schools under common ownership or corporate structure [29], assigns each school a single permanent reference number [29], requires a main school to be a legally established public or private entity [29], and expects consistent school name, campus classification and address across agencies [29], a ready-made reconciliation field. The interstate reciprocity agreement is of limited use for physical campuses, since it applies solely to distance education activity across state lines [37], does not exempt institutions in federally funded programs from compliance [37], and treats a branch campus operating under a main campus's accreditation as not a separate institution [37].

Corporate registries and public filings

Once a registry names an operator, a corporate-registry lookup can confirm the entity. The Delaware Division of Corporations entity search returns, free of charge, details including incorporation or formation date and registered agent [38], but the Division strictly prohibits data mining [38], so bulk matching needs a licensed data route or manual lookups. A registered-agent address alone is weak evidence of common control.

A graded evidence ladder

  • Grade 0, brand only: same public name or domain. A hypothesis; never counted.
  • Grade 1, shared six-digit OPE ID root: same PPA holder. Strong for "same Title IV institution", silent on the ultimate owner.
  • Grade 2, same registry entity across locations: a state record naming one licensee for several sites. Strong for "same operator in that state"; check each jurisdiction.
  • Grade 3, registry or filing naming a controlling parent: the standard for "same ultimate owner". Record the document and its date.
  • Grade 4, event-confirmed: Grade 3 plus the latest ownership-change filing checked after any merger or closure signal.
Figure 02
Shared Title IV institution versus verified owner
Shared Title IV institutionPPA holder
  • A shared six-digit root establishes a shared PPA holder.
  • Different roots can still belong to the same owner.
  • A shared root alone does not establish the ultimate owner.
Verified legal operator and ownerRegistry or filing evidence
  • A state record can establish the same operator across several locations in that state.
  • A registry or filing naming a controlling parent supports the ultimate-owner claim.
  • After a merger or closure signal, check the latest ownership-change filing.

Brand similarity is a hypothesis, never a counted ownership relationship.

A Fictional Crosswalk Worksheet (Hypothetical Example)

Table 3 below is a fictional worksheet for a hypothetical brand, "Example Trade Institute." Every name, number and date is invented to demonstrate the fields; none corresponds to a real school, operator or UnitID. The layout follows the one-to-many logic: one root, several rows, and explicit "unresolved" statuses.

Table 3. Hypothetical one-to-many identifier map (fictional values only)

RowSource URL type and yearJurisdictionUnitIDFull / root OPE IDPPA holderCampus typeFinance-reporting parentPublic brandLegal operator (independently verified)Evidence dateStatus
AState registry, 2026State 190000100999900 / root 009999Example Career Group LLCMain campusSelf (parent for rows B and E)Example Trade InstituteExample Career Group LLC (registry record opened)2026-10-12Verified
BCrosswalk file, 2022-23 vintageState 290000200999902 / root 009999Same PPA root as row ABranch campusChild of row A, partialExample Trade InstituteSame as row A by shared root; State 2 registry check pending2026-10-12Operator hypothesis
CState registry, 2026State 1none (reported inside 900001)00999903 / root 009999Same PPA root as row AAdditional locationNot applicableExample Trade InstituteSame as row A (same licensee at second address)2026-10-12Verified; no separate IPEDS row
DIPEDS directory, 2025-26State 390000401234500 / root 012345Not yet identifiedMain campusSelfExample Trade Institute WestUnresolved: brand similarity only2026-10-12Unresolved affiliation
EIPEDS directory, 2025-26State 190000500999900 / root 009999Same PPA root as row AOnline division (separate UnitID)Child of row A, partialExample Trade OnlineSame as row A by shared root2026-10-12Verified at root level
FState registry, 2026State 4none (no Title IV)noneNot applicableLicensed site, state onlyNot applicableExample Cosmetology AcademyUnresolved: licensee name differs from all other rows2026-10-12Out of IPEDS; unresolved

Rows A, B, C and E share one six-digit root and so one PPA holder: one Title IV institution, with two UnitIDs for the root (rows A and E), consistent with the documented case in which separate campus-based and online UNITIDs sit under one OPEID [22]. Row C is physically a site but has no IPEDS row, consistent with the documented option for a main campus to report combined data for additional locations. Row D shares only a brand, so it is a Grade 0 hypothesis. Row F exists only in a state registry. Counting the six rows as six targets would overstate the universe; a defensible reading is one verified operator across three Title IV-linked rows, two unresolved items and no confirmed relationship between them.

The same worksheet shows what not to sum. Revenue for row B should not be added to row A's if row A is a partial parent, and a campus-level earnings figure for row E would repeat the root-level value. A live version should add a recheck trigger column: any merger, closure, name change, change of PPA holder or ownership notice reopens every row sharing the root, and so does a new crosswalk vintage.

Data Analysis and Evidence

The reporting-unit universe, measured

NCES reported that the number of Title IV institutions in the U.S. and other jurisdictions decreased from 5,918 in 2022-23 to 5,819 in 2023-24 [20]. The 2025-26 provisional IPEDS table lists 5,711 Title IV institutions (1,892 public, 1,721 private nonprofit and 2,098 for-profit), with 1,597 at the less-than-2-year level, of which 1,327 are for-profit (1,290 in the U.S.) [4]. The 2024-25 provisional table shows 5,760 institutions, 2,141 for-profit and 1,350 for-profit less-than-2-year [39].

By simple division, for-profit less-than-2-year institutions were about 63 percent of the for-profit total in both 2024-25 (1,350 of 2,141) and 2025-26 (1,327 of 2,098); these ratios are the author's arithmetic on NCES counts. Of the institutions that were Title IV in 2024-25 but not in 2025-26, 93 of 126 were private for-profit [5], so the population turns over and a frozen list ages quickly. Fall 2024 enrollment at for-profit less-than-2-year institutions was 224,154, about 16.9 percent of for-profit enrollment [40].

The identifier gap, measured by ACE

The ACE paper counted 6,681 unique UnitIDs and 5,253 unique OPEIDs, a gap of 1,428 on its own figures (author's subtraction). It found 345 OPEIDs associated with more than one UnitID, connected to 1,187 UnitIDs in total [3], which averages about 3.4 UnitIDs per such OPEID (author's arithmetic). Approximately 70 percent of those multi-UnitID OPEIDs are for-profit [3], and nearly 96 percent have five or fewer UnitIDs [3]. The paper's data are dated September 2022, so the counts describe a past snapshot; current counts require rerunning the join on a current vintage. The protocol below produces a defensible count:

  • Fix the vintage: record the IPEDS collection year and the crosswalk vintage.
  • Join on the eight-digit OPE ID to retain location detail, then derive the six-digit root.
  • Group by root and list the UnitIDs under each, flagging roots with more than one.
  • Apply parent/child flags from the finance files for the same fiscal year.
  • Attach registry evidence and a ladder grade to each root.
  • Report four labeled counts: locations, UnitIDs, roots, verified operators.

What the evidence does and does not support

The evidence supports three conclusions. Institutions, locations and roots are distinct units whose counts differ for documented structural reasons. Outcome and finance fields can sit at the root or parent level and be repeated across child rows. And the legal operator must come from registries or filings. The evidence does not support a numeric estimate of how many independent operators exist in any vocational universe, and none is offered.

Figure 03
Build a dated, evidence-backed crosswalk
  1. 01Fix the vintage

    Record the IPEDS collection year and the crosswalk vintage.

  2. 02Join at the location level

    Join on the eight-digit OPE ID to retain location detail, then derive the six-digit root.

  3. 03Group by root

    List the UnitIDs under each root and flag roots with more than one.

  4. 04Apply finance reporting flags

    Apply parent/child flags from finance files for the same fiscal year.

  5. 05Attach ownership evidence

    Attach registry evidence and an evidence-ladder grade to each root.

  6. 06Report separate denominators

    Report labeled counts of locations, UnitIDs, roots and verified operators.

Implications and Future Directions

For origination teams, the main effect is on list hygiene. A target list built from IPEDS rows without the bridge will contain duplicates (one operator under several UnitIDs), omissions (satellite sites absent from IPEDS) and false separations (different roots under one owner). Each changes whether a company is a platform candidate, an add-on to another candidate, or already part of a group.

Recheck triggers. A worksheet is a dated artifact. These events should each reopen the affected rows:

  • Ownership change: a reported transfer of ownership interests, since federal rules require reporting of listed changes, generally within 10 days or, for ownership changes of 5 percent to under 25 percent, quarterly.
  • Merger or consolidation: OPE IDs are retained unless the schools apply to merge, so a legal merger may precede an identifier merger.
  • Closure or loss of Title IV status: the 2025-26 release shows institutions leaving the Title IV population.
  • Reporting restructure: a change in how a school reports multisite locations, which NCES says some institutions have made over time.
  • New crosswalk vintage: the Scorecard crosswalk is versioned by year.
  • State action: a new certificate, name change or location change, such as the new certificate Texas requires when ownership changes.

The research gaps are equally clear. The sources reviewed for this report do not publish a national count of operators; the Cheslock note leaves for-profit affiliations unexamined [8]; and the FSA data center pages described in a 2024 announcement were not available for verification as of November 2026, so a public ownership file, if one exists, was not confirmed. Professional counsel should be consulted on regulatory eligibility and change-of-control questions; this report summarizes agency text and is not legal advice.

Frequently Asked Questions (FAQs)

What is an IPEDS UnitID to OPE ID crosswalk, and how do analysts match the two? It is a table linking the IPEDS institution identifier to the eight-digit OPE ID. The Scorecard crosswalk files cover 2000-01 through 2022-23 [2]. Join at eight digits to keep location detail, derive the six-digit root, and expect one-to-many results in both directions.

What is the OPE ID versus UnitID difference? A UnitID names an IPEDS reporting unit; the OPE ID names a Title IV participant or location. They are not a one-to-one match [22].

What is the difference between an eight-digit and a six-digit OPE ID, and the branch code? The six-digit root identifies the institution; the two-digit suffix identifies branches and additional locations [22]. Outcome measures are generally computed at the root.

Where is the College Scorecard OPE ID and UnitID file? On the Scorecard data page, which lists the crosswalk files and was last updated November 2, 2026.

How do IPEDS parent and child institutions work? A parent reports some or all data for a child; see Table 2 for which finance fields are parent-level.

How should private-equity teams map vocational school campuses? Use the denominator bridge in Table 1, label each count with its unit, and grade each parent relationship with the evidence ladder before treating schools as one operator.

Where can multi-campus ownership data for trade schools be found? In state licensing records, which are location-based in some states, in federal ownership-reporting rules, and in corporate registries; none is a single national ownership file.

Conclusion

The IPEDS-to-OPE ID crosswalk is a many-to-many join with documented exceptions. UnitIDs mark IPEDS reporting units, eight-digit OPE IDs mark locations, six-digit roots mark Title IV institutions, and legal operators mark the companies a buyer would approach. The sources opened for this report agree that the first two do not map one-to-one, that outcome data sit at the six-digit level, and that parent-level finance values can duplicate across child rows.

In practice, the defensible workflow keeps the denominators separate, reads each field's reporting level before summing, grades every ownership claim, and rechecks after structural events. It declines to turn a program, campus or UnitID count into an operator count. Peer-reviewed work on administrative units in IPEDS finance data stops short of the for-profit sector, so for-profit and vocational mapping rests on agency documentation and registry evidence, each dated and labeled with its limits. A team that adopts the worksheet fields, the grade ladder and the recheck triggers will be able to say, for any row, which identifier it came from, what level the numbers describe, and how far the ownership claim has been verified. Where a claim stops at Grade 1 or Grade 0, the honest output is an unresolved status, not a guess, and the count of independent operators remains a number to be measured rather than assumed.

External Sources (40)

Published by

Docket Research

Docket provides deal-origination research software and managed research for private equity teams, screening companies against a mandate and showing the evidence behind each finding.

Our platform · Evidence standards · Contact

Get your first 10 targets free

Editorial note

Researched and written by Docket Research with AI-assisted research tools; sources are cited inline and checked before publication. Report an error: https://docket.capital/evidence-standards#corrections.