Arete Healthcare LLC
Independent Practices | Shared Infrastructure | Clinician-Led Governance
Independent Practices | Shared Infrastructure | Clinician-Led Governance
Arete Original Research
Cost, Debt, Earnings, and Graduation Trends Across Postsecondary Credentials — With Focused Analyses of Chiropractic and Physical Therapy
Four figures summarize the paper’s central economic narrative: the relationship between federal debt and early-career earnings, the broader occupational earnings distribution, the interaction of reported earnings progression with debt resolution, and the unusually high self-employment share of chiropractic. Debt-to-earnings is an early-career financing-burden indicator, not a lifetime return-on-investment measure.
Figure 1. Four-Year Earnings and Federal Debt. This figure compares four-year post-graduation earnings with evaluated federal student debt across bachelor’s, master’s, doctoral, and professional-practice fields of study. Each point represents a field-of-study/credential combination with at least 10 borrowers; point color denotes credential level, and bubble size represents the number of individuals contributing to the four-year earnings estimate. Vertical and horizontal reference lines identify the median four-year earnings and median federal debt among plotted observations. The shaded area identifies fields in which evaluated federal debt is lower than four-year earnings, while observations above the diagonal have debt exceeding four-year earnings. Selected professional fields are labeled to illustrate differences in early-career economic burden. Chiropractic occupies a distinct high-debt/lower-earnings position relative to several other professional health fields, whereas medicine, dentistry, optometry, pharmacy, and law generally combine substantial debt with materially higher four-year earnings. Visualization was human generated.
Data source: U.S. Department of Education College Scorecard Field of Study data. Earnings measures are derived from federal administrative earnings records linked to educational records. Federal debt reflects the College Scorecard field-of-study debt construct for evaluated borrowers and should not be interpreted as total lifetime educational cost or necessarily as all undergraduate plus graduate borrowing. Earnings and debt populations may differ because of reporting, suppression, and eligibility requirements. Four-year earnings are measured among individuals meeting the Scorecard earnings-population criteria and should not be interpreted as the mean earnings of all graduates. The figure is descriptive and does not establish that debt level causes subsequent earnings or economic outcomes.
Figure 2. Health Professions Median, 25th, and 75th Percentile Earnings — 2025 BLS. This figure places the early-career College Scorecard findings within the broader occupational earnings environment by presenting the 25th percentile, median, and 75th percentile annual wages for selected healthcare occupations. The central marker represents the occupational median wage and the vertical range represents the 25th-to-75th percentile wage distribution. Chiropractors, acupuncturists, physical therapists, and selected comparison occupations are highlighted to facilitate interpretation. The figure demonstrates substantial variation both across healthcare professions and within individual occupations; occupations with similar median wages can have markedly different earnings distributions. These data provide context for longer-term occupational earning potential but do not represent the earnings of recent graduates. Visualization was human generated.
Data source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, 2025. OEWS estimates represent wage-and-salary workers and exclude self-employed workers and owners of unincorporated businesses. This exclusion is particularly important for professions with high self-employment rates, including chiropractic and acupuncture. BLS occupational categories also do not map perfectly to educational fields or CIP codes, and the estimates include workers at all career stages rather than graduates at a fixed number of years after degree graduation. Accordingly, these values should not be directly substituted for College Scorecard post-graduation earnings or interpreted as longitudinal earnings trajectories.
Figure 3. Reported Earnings Progression and Student-Loan Debt Repayment Status. This figure examines the relationship between reported early-career earnings progression and observed federal student-loan repayment performance. The horizontal axis is the ratio of reported four-year earnings to reported one-year earnings for each field-of-study/credential combination; values above 1.0 indicate that reported four-year earnings exceed reported one-year earnings. The vertical axis is the estimated proportion of borrowers with positive amortization, defined in this study as borrowers classified by the College Scorecard Borrower-Based Repayment Rate as either Making Progress — outstanding balances below original balances while in repayment — or Paid in Full. Point color denotes credential level and bubble size represents the number of individuals contributing to the four-year earnings estimate. Median reference lines divide the figure into relative earnings-progression and debt-resolution regions. Chiropractic appears in the lower-left portion of the distribution, combining relatively limited reported earnings progression with a comparatively low estimated share of borrowers reducing or fully repaying principal. Medicine and dentistry illustrate that similarly large educational debt can have different repayment implications when post-graduation earnings rise more rapidly. Visualization was human generated.
Data source: U.S. Department of Education College Scorecard Field of Study earnings and Borrower-Based Repayment Rate data. Positive amortization is an Arete-derived measure equal to the sum of the Scorecard Making Progress and Paid in Full categories. Because Scorecard privacy protection frequently reports BBRR proportions as intervals rather than exact values, plotted percentages use borrower-weighted interval midpoint estimates and therefore are analytic estimates rather than official exact percentages. BBRR categories are hierarchical: borrowers classified in default, delinquency, forbearance, or deferment are not subsequently classified as making or not making progress. Repayment cohorts overlapping the COVID-19 period may also have been affected by pandemic-related administrative forbearance. In addition, the one-year and four-year Scorecard earnings measures in the Most Recent file are not necessarily observations from the same graduating cohort; the Y4/Y1 ratio should therefore be interpreted as a reported earnings-progression proxy, not as within-person longitudinal earnings growth. Suppression and differing data availability reduce the number of fields represented in the repayment analysis.
Figure 4. Health Professions Self-Employed Percent — 2025 BLS. This figure compares the estimated percentage of workers who are self-employed across selected healthcare professions. Chiropractic has one of the highest self-employment shares among the healthcare occupations examined, with approximately 36% of chiropractors classified as self-employed; only acupuncture is higher among the selected occupations. By comparison, self-employment is substantially less common among dentists, veterinarians, optometrists, physical therapists, physicians, nurse practitioners, and pharmacists. The figure provides important context for interpreting both graduate earnings and occupational wage estimates: graduates entering professions characterized by substantial self-employment may depend more heavily on successful practice formation, revenue growth, and operating-cost management than graduates entering predominantly employed professions. Visualization was human generated.
Data source: U.S. Bureau of Labor Statistics Employment Projections occupational characteristics, 2025 base-year estimates. Self-employment percentages describe the occupational workforce and are not limited to recent graduates. Occupational classifications do not correspond perfectly to educational CIP fields, and the measure does not distinguish among sole proprietorships, partnerships, incorporated practice ownership, independent contracting arrangements, or other forms of business ownership. Importantly, the self-employed population shown here is generally outside the wage-and-salary population represented in the BLS OEWS earnings distributions in Figure 2; consequently, the limitation of OEWS wage data is disproportionately relevant to chiropractic and other highly self-employed professions. The figure describes workforce structure and should not be interpreted as evidence that self-employment itself produces higher or lower earnings.
This Arete Original Research paper was developed in response to an invitation to participate in a Faculty American Chiropractic Association (FACA) roundtable to discuss “The ROI of Chiropractic Education.”
Previous Arete Original Research has explored several unique attributes of the chiropractic profession: The Organizational Outlier, The Digital Outlier, and The Funding Outlier. These Original Research papers demonstrated that the chiropractic profession is an outlier organizationally, technologically, and economically. A Workforce in Transition and The Pipeline in Transition demonstrated how the chiropractic workforce is evolving from primarily white male to a more diverse profession. Chiropractors were distinctive for being one of only two specialties with fewer clinicians graduating in 2021-25 than 1995-99, with chiropractors having the largest reduction of 20.4%.
This Original Research paper adds another layer to the evolving workforce story, and may help explain the apparent contraction in the chiropractic workforce. Medicine and other professional degrees convert a high initial debt burden into a favorable cost-benefit equation through manageable early-career earnings that accelerated during the first five years. Chiropractic largely does not.
The historical broad access to federal graduate/professional borrowing could reduce chiropractic students’ near-term sensitivity to price and early-career economics, permitting educational demand to persist before those economics are fully observable.
Chiropractors have an unusually entrepreneurial employment structure, being more likely to be self-employed than all other health professions except licensed acupuncturists. This means early-career economic success may depend more heavily on practice formation, revenue growth, and cost management than in predominantly employed health professions.
While this analysis included institution-level analyses, institution names were masked because the purpose was to characterize variation rather than rank individual institutions. Given the economic story revealed in the paper, a surprising finding was that higher cost chiropractic programs tended to grow over the past decade both in absolute terms and relative to lower cost programs. This finding was not uniform among institutions, and the trend was weaker over the shorter, more recent horizon. We will leave it to others to speculate about why this might be, and whether this trend may continue.
Finally, this Original Research paper provides further insights into the Arete “Why” – supporting practices navigating complex economics to express their full potential to help patients through fantastic clinical care and connection.
Modern, efficient, effective and low-cost shared MSO platform to reduce the operational costs of practice, and maximize collections
Secure contracts where practice reimbursement is aligned with value
Support academic institution clinic systems and curriculum to model for students in training what it takes to succeed in practice
Develop and distribute new practitioner readiness curriculum and ongoing continuing education through Friday15
Our healthcare system needs the high-value care provided by chiropractors and we could not be more optimistic about the future.
The analysis combines College Scorecard debt, borrower repayment status, and post-graduation earnings with longitudinal IPEDS graduations and cost data. The table below shows the reported Scorecard earnings series from one through five years after graduation alongside debt, debt-to-four-year-earnings burden, and graduation growth. A companion debt-resolution snapshot adds Debt/Y1, Debt/Y4, four-year Borrower-Based Repayment Rate (BBRR4) status, and graduation trends for selected professional fields. For the four credential-level populations, earnings are rolled up as worker-count-weighted averages of institution-program medians, so larger reporting populations contribute proportionally rather than every field receiving equal weight. The primary cross-field burden measure remains cumulative evaluated-institution debt divided by national median earnings four years after graduation. Graduation trends use the 2018-19 to 2024-25 short-term CAGR and the 2014-15 to 2024-25 long-term CAGR, with 2019-20 retained as a COVID-overlap sensitivity.
| Population | Earnings: Y1 / Y2* / Y3* / Y4 / Y5 | Debt | Debt / earnings | Graduation CAGR: short / long |
|---|---|---|---|---|
| Bachelor's fields | $49k / $42k* / $49k* / $68k / $65k | $21,469 | 0.33x | -1.3% / -0.1% |
| Master's fields | $73k / $63k* / $68k* / $89k / $86k | $42,671 | 0.52x | +0.1% / +1.3% |
| Doctoral (research/other) | $90k / $79k* / $87k* / $110k / $112k | $76,368 | 0.89x | +1.3% / +2.4% |
| Professional-practice fields | $84k / $77k* / $91k* / $125k / $137k | $115,220 | 1.10x | +1.7% / +1.7% |
| Chiropractic | $49k / $39k* / $44k* / $61k / $60k | $196,057 | 3.21x | +2.5% / +1.7% |
| PT-dominant 51.23 rehabilitation field | $75k / $70k* / $73k* / $82k / $83k | $109,155 | 1.34x | +4.8% / +5.2% |
| Alternative/complementary medicine | $33k / $23k* / $32k* / $42k / $44k | $143,279 | 3.44x | -4.6% / +3.8% |
| Medicine | $65k / $58k* / $68k* / $112k / $173k | $183,753 | 1.63x | +7.7% / +5.2% |
| Dentistry | $117k / $109k* / $129k* / $171k / $157k | $264,439 | 1.55x | +1.7% / +1.8% |
| Optometry | $98k / $95k* / $108k* / $132k / $119k | $182,745 | 1.39x | -0.5% / +0.8% |
| Pharmacy | $101k / $118k* / $125k* / $138k / $143k | $141,199 | 1.02x | -4.8% / -2.6% |
| Veterinary medicine | $101k / $79k* / —* / $131k / $121k | $178,661 | 1.36x | — / — |
| Law | $87k / $74k* / $89k* / $108k / $121k | $121,285 | 1.13x | +1.0% / -1.0% |
*Earnings are displayed Y1 / Y2* / Y3* / Y4 / Y5. The asterisk marks legacy Scorecard Y2/Y3 metrics from earlier reporting vintages; the three-year measure used different eligibility rules and included nonworkers. Credential rows use worker-count-weighted averages of OPEID6-deduped institution-program medians at each measurement point, using all reporting fields in the credential. Field rows continue to use weighted averages of institution-program medians for Y1/Y2/Y3/Y5, while Y4 is the Scorecard national field median used in the field burden ratio. Because worker populations, cohorts, eligibility rules, and inflation reference years differ by measure, these observations should not be interpreted as the earnings trajectory of one longitudinal cohort. Credential-row debt and debt/earnings values remain median-across-field burden summaries and are analytically separate from the displayed worker-weighted earnings series. CIP 51.23 is PT-dominant but mixed; exact six-digit DPT graduation trends are used where stated.
| Field | Debt | Y1 → Y4 earnings | Debt/Y1 → Debt/Y4 | BBRR4: FBR / NOPROG / +amort. | BBRR4 N | Graduation CAGR: short / long |
|---|---|---|---|---|---|---|
| Chiropractic | $196,057 | $49,051 → $61,163 | 4.00x → 3.21x | 74% / 17% / 13% | 2,272 | +2.5% / +1.7% |
| Medicine | $183,753 | $65,207 → $112,460 | 2.82x → 1.63x | 47% / 26% / 26% | 21,940 | +7.7% / +5.2% |
| PT-dominant rehab* | $109,155 | $74,773 → $81,641 | 1.46x → 1.34x | 44% / 20% / 35% | 3,794 | +4.8% / +5.2% |
| Dentistry | $264,439 | $116,891 → $170,812 | 2.26x → 1.55x | 37% / 18% / 44% | 6,500 | +1.7% / +1.8% |
| Pharmacy | $141,199 | $101,042 → $138,348 | 1.40x → 1.02x | 36% / 17% / 45% | 14,866 | -4.8% / -2.6% |
Debt-resolution snapshot. BBRR4 values are borrower-weighted midpoints of unsuppressed/coarsened program ranges; coverage differs by status and is reported in Appendix A. FBR = forbearance; NOPROG = not making progress; +amort. = making progress + paid in full. Forbearance is especially COVID-sensitive. *CIP 51.23 is PT-dominant but mixed.
Interpretation. The credential-level financial-burden gradient is clear, and the pooled negative association with graduation growth is robust to both short- and long-term pre-COVID baselines. Chiropractic remains an important exception: national graduations increased over both horizons, and institution-level burden does not show the pooled negative pattern.
Background: Postsecondary cost-benefit comparisons are often built from sticker tuition or mature-career wage estimates that do not align by credential, field, cohort, or institution. This study integrates federal administrative and survey-derived datasets to characterize early-career financial burden across bachelor’s, master’s, doctoral, and professional-practice education, with focused analyses of chiropractic and physical therapy.
Methods: College Scorecard field-of-study data supplied cumulative Stafford plus Graduate PLUS debt originated at the evaluated institution, borrower-based repayment status, national four-year earnings, and additional post-graduation earnings observations at years 1 through 5. BBRR4 categories were analyzed as hierarchical borrower statuses four years after entry into repayment; privacy-coarsened ranges were retained with midpoint sensitivity estimates and explicit coverage. Credential-level Y1-Y5 earnings were summarized as worker-count-weighted averages of institution-program medians using metric-specific worker counts; Y2/Y3 were retained as legacy descriptive measures rather than treated as a same-cohort trajectory. IPEDS supplied 2011-12 through 2024-25 graduations and longitudinal tuition/cost data. Graduation trends used pre-COVID short-term (2018-19 to 2024-25) and long-term (2014-15 to 2024-25) CAGRs, with additional sensitivity analyses. BLS and IRS SOI supplied workforce-structure context.
Results: Median debt/earnings burden rose from 0.33x for bachelor’s fields to 1.10x for professional-practice fields. Chiropractic measured 3.21x, compared with 1.34x for the PT-dominant 51.23 rehabilitation field. Chiropractic and medicine had similar evaluated federal debt (~$196,057 vs ~$183,753), but reported Y1/Y4 earnings differed (~$49,051/$61,163 vs ~$65,207/$112,460), leaving Debt/Y4 at 3.21x versus 1.63x. Among visible/coarsened BBRR4 data, chiropractic showed an estimated 74% forbearance midpoint, 17% not-making-progress midpoint, and 13% positive-amortization midpoint, versus 47%, 26%, and 26% for medicine; COVID-era administrative forbearance limits interpretation. Across 591 eligible field-by-credential observations, +1.0x higher burden was associated with 3.60 percentage points lower annual short-term graduation CAGR (p<0.001), with consistent long-term/sensitivity direction. Chiropractic graduations nevertheless increased 2.5% annually short-term and 1.7% annually long-term.
Conclusion: Early-career financing burden and debt-resolution profiles differ markedly by credential and professional field. Chiropractic is simultaneously a financial-burden outlier, a repayment-resolution concern, a workforce-structure outlier, and a graduation-response exception. Growing graduations despite unfavorable graduate economics create a financing-demand paradox that is consistent with several mechanisms, including delayed price sensitivity, but the available data do not establish federal loan availability as the cause or demonstrate an imminent tipping point.
This exploratory observational study asked three questions: (1) how early-career debt burden differs across credential levels and fields; (2) how healthcare professional fields compare, particularly chiropractic and physical therapy; and (3) whether less favorable financial burden is associated with declining graduations. A third, institution-level analysis examined chiropractic programs separately.
| Term | Definition |
|---|---|
| CAGR | Compound annual growth rate. Primary short-term CAGR links 2018-19 to 2024-25: (C2024-25 / C2018-19)^(1/6) - 1. Primary long-term CAGR links 2014-15 to 2024-25: (C2024-25 / C2014-15)^(1/10) - 1. The 2019-20 to 2024-25 five-year CAGR is retained as a COVID-overlap sensitivity. |
| CBA | Cost-benefit analysis. The current paper uses an early-career debt/earnings burden proxy; it does not claim a complete lifetime cost-benefit or net-present-value estimate. |
| CIP | Classification of Instructional Programs, the NCES taxonomy for fields of study. Scorecard field outcomes are generally four-digit CIP; IPEDS graduations provide six-digit detail. |
| Debt / 4-year earnings | Mean cumulative evaluated-institution debt divided by national median earnings four years after graduation. Lower values are more favorable. |
| Evaluated-institution debt | College Scorecard Stafford plus Graduate PLUS debt originated at the evaluated institution for the same academic level; it is not total lifetime educational debt. |
| HC3 | A heteroskedasticity-consistent standard-error estimator that is robust to nonconstant residual variance and gives additional protection against leverage in finite samples. |
| IPEDS | Integrated Postsecondary Education Data System, NCES institutional census data used here for graduations, program classification, and tuition/cost measures. |
| NPSAS | National Postsecondary Student Aid Study. NPSAS:20 was evaluated but DC-specific estimates were suppressed because fewer than 30 valid sample cases supported them. |
| OEWS | BLS Occupational Employment and Wage Statistics. OEWS wage estimates exclude self-employed workers. |
| PSEO | Census Postsecondary Employment Outcomes, used as an earnings sensitivity source where participating institutions are represented. |
| Self-employed share | BLS Employment Projections estimate of the percentage of workers in an occupation who are self-employed. This is the study's prevalence measure for workforce structure. |
| SOI | IRS Statistics of Income. The 2023 nonfarm sole-proprietorship tables summarize Schedule C business returns by industry. |
| Sole-proprietorship return | An IRS Schedule C business tax return. It is not equivalent to one graduate or one worker and cannot be divided by BLS employment to estimate sole-proprietor prevalence. |
| Professional-practice degree | IPEDS/Scorecard credential category for degrees preparing graduates for professional practice; distinguished from doctoral research/scholarship and other doctorates. |
| Field-level burden ratio | For a field/credential, pooled mean evaluated-institution debt divided by the field’s national median earnings four years after graduation. This is the primary field-level burden measure used in cross-field analyses. |
| Program-level burden ratio | For an institution-program observation, institution-specific evaluated-institution mean debt divided by institution-specific median earnings four years after graduation. Within-field program P25/median/P75 values use only unsuppressed numeric observations and are not interchangeable with the field-level pooled ratio. |
| Scorecard earnings series | Reported post-graduation earnings observations at years 1 through 5. Y2 and Y3 are legacy metrics from earlier Scorecard vintages; Y3 used different eligibility rules and included nonworkers. Field summaries use weighted program medians for Y1/Y2/Y3/Y5 and the national field median for Y4. Credential summaries use worker-count-weighted averages of institution-program medians for Y1-Y5. The five observations are descriptive and are not a same-cohort longitudinal series. |
| Worker-count-weighted program-median earnings | A constructed aggregate equal to sum(institution-program median earnings x the corresponding Scorecard worker count) divided by sum(worker counts). For credential-level summaries, this gives larger reporting populations proportional influence and avoids treating every field as equally sized. It is not an individual-level mean earnings estimate. |
| BBRR / BBRR4 | Borrower-Based Repayment Rate; BBRR4 classifies federal borrowers four years after entering repayment into a hierarchical, mutually exclusive status. |
| Not making progress | BBRR status for borrowers making regular payments whose total outstanding balance still exceeds original balances, after excluding default, delinquency, forbearance, and deferment. |
| Positive amortization | Study-defined descriptive construct equal to BBRR making progress + paid in full; not an official Scorecard metric. |
| BBRR midpoint | Study sensitivity estimate for a privacy-coarsened Scorecard interval; not an official Scorecard point estimate. |
| Reported burden compression | Debt/Y1 minus Debt/Y4. Descriptive cross-cohort measure of how much the reported debt/earnings ratio differs between the Y1 and Y4 earnings observations; not within-person loan amortization. |
| Graduation-financing paradox | Coexistence of unfavorable graduate economics with stable or growing graduations. The term describes the observation; it does not establish that federal financing caused the growth or that a tipping point is imminent. |
Source URLs and file inventory are provided in the References and accompanying analytic workbook.
| Source | Primary use | Role in analysis |
|---|---|---|
| College Scorecard | Field/credential debt, repayment, Y1-Y5 post-graduation earnings observations; institution outcomes | Core |
| IPEDS Graduations | 2011-12 to 2024-25 graduations; six-digit CIP resolution | Core |
| IPEDS Cost / Institutional Characteristics | Historical tuition/cost; selected professional-practice tuition | Core/context |
| BLS OEWS & projections | Occupational wages, projections, self-employment context | Context |
| Census PSEO | Year 1/5/10 earnings where participating institutions are available | Sensitivity |
| IRS SOI | Sole-proprietorship chiropractic practice economics | Context only |
| FAFSA earnings file | Independent file-level check of Scorecard-derived undergraduate earnings | Validation |
| NPSAS:20 PowerStats | Evaluated for debt/aid; DC-specific cells suppressed | De-scoped for DC quantitation |
Core = essential to the primary analysis. Context = provides interpretive triangulation but does not enter the primary burden or regression measures. Sensitivity/Validation = used to check, not replace, the primary Scorecard-based results.
The College Scorecard field-of-study files aggregate programs at the four-digit Classification of Instructional Programs (CIP) level and credential level. Where the source file repeated the same OPEID6 x CIP x credential outcome across multiple UNITIDs, the repeated observations were treated as one institution-program record before aggregation. The primary burden analysis used national median four-year post-graduation earnings and cumulative evaluated-institution debt. To make the early-career earnings pattern visible, the study also retained Scorecard earnings observations at years 1 through 5. For field summaries, Y1/Y2/Y3/Y5 were calculated as worker-count-weighted averages of institution-program medians available in the Most Recent field file; Y4 used the Scorecard national field median. For credential-level earnings summaries, the institution-program median series was rolled up using the corresponding metric-specific worker counts: sum(program median × worker count) / sum(worker count) across all reporting fields/programs in the credential. This preserves population weighting rather than assigning equal weight to each field. The credential aggregation remains an average of program medians, not an individual-level mean. Y2 and Y3 are legacy metrics from prior reporting vintages, and Y3 included nonworkers under a different eligibility rule. Because the Most Recent file combines different cohorts, eligibility rules, worker populations, and inflation reference years, the Y1-Y5 observations are presented descriptively and are not interpreted as within-person or same-cohort earnings growth. The debt measure represents Stafford plus Graduate PLUS debt originated at the institution being evaluated for the same academic level; it is a financing measure, not total educational cost, and does not add undergraduate debt to a graduate/professional debt observation.
Scorecard credential level 6 combines IPEDS research/scholarship and other doctorates. Accordingly, this paper uses the label doctoral (research/other) and does not describe that category as PhD. Credential level 7 is treated as professional practice / first professional.
IPEDS first-major graduations were normalized from 2011-12 through 2024-25 by institution, six-digit CIP, and award level. Historical academic-year student-charge files were used through 2023; the separate IPEDS Cost component was used beginning in 2024. Selected professional-practice tuition is directly reported for chiropractic and several other professions. General graduate tuition is not substituted for professional-practice programs when the IPEDS reporting framework excludes those programs. [2,3]
For chiropractic institution cost context, the analysis calculated a trailing four-academic-year sum of institution-specific reported chiropractic tuition and required fees. This is a nominal published-tuition proxy and excludes living expenses, grants, financing interest, opportunity cost, and other educational expenses. Source-level quality flags were retained rather than silently correcting implausible observations.
College Scorecard reports professional-practice earnings and debt for four-digit CIP 51.23, Rehabilitation and Therapeutic Professions, rather than six-digit CIP 51.2308 Physical Therapy/Therapist. IPEDS 2024-25 data show 12,223 DPT graduations among 15,753 professional-practice graduations in 51.23 (77.6%). The paper therefore describes the Scorecard observation as PT-dominant but not PT-specific. Exact DPT graduation trends use six-digit CIP 51.2308.
The primary early-career burden indicator was mean cumulative evaluated-institution debt divided by national median earnings four years after graduation. This ratio was selected because it is available at a common field-by-credential level across a large number of programs. It is a financing-burden proxy, not a complete lifetime cost-benefit measure. For selected professional fields, Scorecard coverage was also summarized as unique OPEID6 x field x credential institution-program records. A separate within-field distribution used institution-specific debt divided by institution-specific four-year earnings among unsuppressed records; its P25, median, and P75 values are reported as program-level dispersion and are not substituted for the field-level pooled burden ratio.
Graduation trends were evaluated over two primary horizons. Short-term CAGR used 2018-19, the last full academic year preceding COVID-19 disruption, as the baseline and 2024-25 as the endpoint: (C2024-25 / C2018-19)^(1/6) - 1. Long-term CAGR used 2014-15 as the baseline and 2024-25 as the endpoint: (C2024-25 / C2014-15)^(1/10) - 1. The original 2019-20 to 2024-25 five-year CAGR was retained as a COVID-overlap sensitivity. CAGR was not calculated when the relevant baseline count was zero. A declining-program sensitivity indicator was defined as annualized CAGR below -2%.
College Scorecard BBRR4 classifies federal loan borrowers four years (4 × 365 days) after entering repayment into eight mutually exclusive, hierarchical statuses: default, delinquent, forbearance, deferment, not making progress, making progress, paid in full, and discharged. Because a borrower is assigned to the first applicable status, not-making-progress excludes borrowers already classified as defaulted, delinquent, in forbearance, or in deferment. [1]
BBRR values are privacy protected. Unsuppressed values may be exact, rounded, or coarsened into ranges. The analysis preserved source ranges and calculated interval midpoints only as sensitivity estimates: a-b used (a+b)/2; ≤x used x/2 with bounds 0 to x; ≥0.80 used 0.90 with bounds 0.80 to 1.00. Program values were weighted by the BBRR4 borrower cohort count after OPEID6 x four-digit CIP x credential deduplication. Status-specific borrower coverage is reported because suppression differs by category. Positive amortization was defined for this study as making progress + paid in full; it is not an official Scorecard metric.
The debt-resolution profile combines evaluated-institution debt, reported Y1 and Y4 earnings, Debt/Y1, Debt/Y4, the reported Y4/Y1 earnings multiple, BBRR4 status, and IPEDS graduation growth. These components do not follow the same borrower cohort. Reported burden compression (Debt/Y1 minus Debt/Y4) is therefore descriptive rather than a within-person amortization measure. Scorecard also cautions that repayment measurements overlapping award years 2018-19 and 2019-20 may have been affected by administrative forbearance and other pandemic-related borrower-status changes; forbearance is not interpreted as ordinary repayment distress. [1]
The short-term primary sample included 591 field-by-credential observations with numeric Scorecard debt and four-year earnings, numeric IPEDS short-term CAGR, and at least 30 graduations in both 2018-19 and 2024-25. The long-term primary sample included 578 observations meeting the same criteria in 2014-15 and 2024-25. Ordinary least squares models used HC3 robust standard errors and adjusted for credential fixed effects, log baseline graduations for the relevant horizon, and healthcare status. Sensitivity analyses restricted burden to ≤2x, required at least 100 baseline/current graduations, retained the 2019-20 COVID-overlap baseline, and estimated a log-linear annual graduation trend across 2014-15 through 2024-25 when at least eight annual observations were positive. These analyses are associational; the current burden measures overlap with the historical graduation periods and are not interpreted causally.
IPEDS identified 18 current chiropractic institutions. College Scorecard suppression left 10 institutions with numeric professional-practice debt and four-year earnings for the institution-level burden analysis. Spearman and Pearson correlations compared institution debt/earnings and tuition/earnings ratios with short-term (2018-19 to 2024-25), long-term (2014-15 to 2024-25), and COVID-overlap sensitivity graduation CAGRs. These analyses were designated exploratory because of the small sample, suppressed data, potential institutional changes, overlap between current burden measures and historical graduation windows, and lack of causal identification.
BLS OEWS and Employment Projections were used as mature occupational-wage and workforce-structure context. OEWS wage estimates exclude self-employed workers, while the Employment Projections data separately estimate the share of workers who are self-employed. IRS SOI nonfarm sole-proprietorship tables were used to characterize Schedule C business-return economics and were not interpreted as graduate earnings, worker counts, or a denominator for sole-proprietor prevalence. Census PSEO was retained as a sensitivity source where institutions participate, but only three of 18 current chiropractic institutions were represented in the supplied professional-practice chiropractic earnings data and PSEO primarily reflects unemployment-insurance-covered employment. FAFSA earnings data independently matched corresponding College Scorecard undergraduate earnings observations and were treated as validation rather than a second outcome source. [4-6]
NPSAS:20 Graduate PowerStats was evaluated because it explicitly identifies Chiropractic (DC, DCM) and includes borrowing, amount owed, tuition, and aid measures. In the public PowerStats output, chiropractic-specific cells returned the NCES double-dagger suppression symbol across the requested financial measures because fewer than 30 valid sample cases supported the estimates. Public PowerStats also does not expose institution identifiers for institution-level analysis. Restricted-use access would not increase the underlying chiropractic sample. NPSAS:20 was therefore de-scoped from chiropractic-specific quantitative estimates, while remaining an evaluated source for possible broad degree-level context. Chiropractic should be re-tested when NPSAS:24 public data become available. [7]
The accompanying analytic workbook retains source inventory, normalized field and institution tables, source-specific QA decisions, formulas for derived burden measures, and recorded statistical output. Retained Python scripts document the integrated analytic build and the dual-horizon graduation update used for this version.
The median debt-to-four-year-earnings ratio increased monotonically across credential levels: 0.33x for bachelor’s fields, 0.52x for master’s fields, 0.89x for doctoral (research/other) fields, and 1.10x for professional-practice fields. Median four-year earnings also increased with credential level, but cumulative evaluated-institution debt increased more rapidly. Figure 1 summarizes this gradient.
The broader Scorecard earnings series adds an important time dimension to the burden measure. Across professional-practice programs represented in Scorecard, the worker-count-weighted average of institution-program medians was approximately $83,740 at Y1, $125,251 at Y4, and $137,255 at Y5. The analogous worker-weighted Y1/Y4/Y5 values were approximately $48,746/$67,893/$64,650 for bachelor’s programs, $72,812/$89,120/$85,552 for master’s programs, and $89,514/$110,109/$112,346 for doctoral (research/other) programs. Chiropractic reported approximately $49,051 at Y1, $61,163 at Y4, and $60,133 at Y5, compared with $74,773, $81,641, and $82,754 for the PT-dominant 51.23 rehabilitation field. Legacy Y2/Y3 observations are reported in At a Glance and Appendix A but are not interpreted as within-cohort earnings changes because their cohorts, worker populations, and eligibility rules differ.
Among 16 professional-practice fields with complete debt and four-year earnings data, the distribution was wide. Alternative and Complementary Medicine and Medical Systems had the highest burden ratio at 3.44x. Chiropractic was second at 3.21x, based on mean evaluated-institution debt of $196,057 and median four-year earnings of $61,163. Medicine measured 1.63x, dentistry 1.55x, optometry 1.39x, veterinary medicine 1.36x, the PT-dominant 51.23 rehabilitation field 1.34x, law 1.13x, and pharmacy 1.02x. Nursing and allied-health diagnostic/intervention fields were below 1.0x. Figure 1 presents the debt-versus-four-year-earnings landscape; Appendix A additionally reports Scorecard program coverage and within-field institution-level burden dispersion for selected health professional fields.
Similar debt levels produced very different reported early-career burden profiles. Chiropractic mean evaluated-institution debt was $196,057 versus $183,753 for Medicine (CIP 51.12). Chiropractic reported approximately $49,051 at Y1 and $61,163 at Y4, compared with $65,207 and $112,460 for medicine. Debt/Y1 was therefore approximately 4.00x for chiropractic versus 2.82x for medicine, and Debt/Y4 was 3.21x versus 1.63x. The reported Y4/Y1 earnings multiple was 1.25x for chiropractic versus 1.72x for medicine. These comparisons are descriptive because the Scorecard Y1 and Y4 measures do not represent the same longitudinal cohort.
| Field | Debt | Y1 / Y4 | Debt/Y1 → Y4 | BBRR4 FBR | BBRR4 NOPROG | BBRR4 +amort. | Short / long CAGR |
|---|---|---|---|---|---|---|---|
| Chiropractic | $196,057 | $49,051 / $61,163 | 4.00x → 3.21x | 74% | 17% | 13% | +2.5% / +1.7% |
| Medicine | $183,753 | $65,207 / $112,460 | 2.82x → 1.63x | 47% | 26% | 26% | +7.7% / +5.2% |
| PT-dominant rehab* | $109,155 | $74,773 / $81,641 | 1.46x → 1.34x | 44% | 20% | 35% | +4.8% / +5.2% |
| Dentistry | $264,439 | $116,891 / $170,812 | 2.26x → 1.55x | 37% | 18% | 44% | +1.7% / +1.8% |
| Pharmacy | $141,199 | $101,042 / $138,348 | 1.40x → 1.02x | 36% | 17% | 45% | -4.8% / -2.6% |
BBRR4 percentages are midpoint sensitivity estimates of unsuppressed/coarsened program ranges weighted by BBRR4 borrower N; coverage differs by status. FBR = forbearance; NOPROG = not making progress; +amort. = making progress + paid in full. Forbearance is especially sensitive to COVID-era administrative forbearance.
The a priori hypothesis that chiropractic would have an unusually high raw BBRR4 not-making-progress rate was only partly supported. Among borrower-weighted visible/coarsened values, the chiropractic not-making-progress midpoint was approximately 17% (87% borrower coverage), compared with 26% for medicine, 20% for the PT-dominant rehabilitation field, 18% for dentistry, and 17% for pharmacy. Chiropractic was more distinctive for very high forbearance (~74% midpoint; 95% coverage) and low positive amortization (~13% midpoint; 76% coverage), compared with medicine (~47% and ~26%), the PT-dominant field (~44% and ~35%), dentistry (~37% and ~44%), and pharmacy (~36% and ~45%).
Among the subset of programs where not making progress, making progress, and paid in full were all observable, the midpoint share not making progress among these active repayment outcomes was approximately 58% for chiropractic versus 51% for medicine. Privacy-coarsened intervals are broad, however, and this conditional construct excludes borrowers in higher-priority BBRR statuses. The repayment data therefore support concern about weak debt resolution in chiropractic but do not support a simple claim that raw not-making-progress is uniquely elevated.
Across 591 short-term eligible observations, each +1.0x increase in debt relative to four-year earnings was associated with a 0.0360 lower annual graduation CAGR, equivalent to 3.60 percentage points per year (95% CI, -5.63 to -1.57; p<0.001; R²=0.049). The long-term model included 578 observations and estimated a 0.0279 lower annual CAGR per +1.0x burden (95% CI, -4.57 to -1.00; p=0.002; R²=0.063). The original 2019-20 baseline produced a similar coefficient (-0.0277; p=0.003), and the full-series log-linear sensitivity remained negative (-0.0255; p=0.006). The direction of the adjusted pooled association therefore did not depend on selecting 2019-20 as the baseline year.
| Model | N | Debt/earnings coefficient | p value | R² / OR |
|---|---|---|---|---|
| Primary short-term HC3 OLS | 591 | -0.0360 | <0.001 | R² 0.049 |
| Primary long-term HC3 OLS | 578 | -0.0279 | 0.002 | R² 0.063 |
| COVID-overlap sensitivity | 596 | -0.0277 | 0.003 | R² 0.044 |
| Full-series log-linear sensitivity | 578 | -0.0255 | 0.006 | R² 0.061 |
OLS coefficients are changes in annualized graduation growth for a +1.0x change in debt/earnings burden. Short-term and long-term models use their respective pre-COVID baselines; the COVID-overlap and full-series log-linear models are sensitivity analyses.
Unadjusted Spearman correlations remained small in the pooled samples, underscoring the importance of credential-level differences and baseline program size. Within-credential burden quartiles showed a clearer gradient: median short-term CAGR declined from +1.43% in Q1 to -0.69% in Q4, while median long-term CAGR declined from +2.11% in Q1 to -0.50% in Q4.
Repayment status did not provide a stronger contemporaneous explanation of professional-program graduation trends in exploratory analyses. Across professional-practice fields, higher BBRR4 not-making-progress and active no-progress midpoints were directionally associated with weaker short- and long-term graduation growth, but correlations were not statistically significant. Debt/Y4 retained a clearer inverse association with short-term graduation growth. Given privacy coarsening, COVID effects, cross-cohort timing, and small field counts, these BBRR/graduation comparisons are hypothesis-generating rather than confirmatory.
Chiropractic combined a high 3.21x debt/earnings ratio with increasing national graduations over both primary horizons. Exact DC graduations rose from 2,608 in 2018-19 to 3,022 in 2024-25, a +2.5% annual short-term CAGR, and from 2,544 in 2014-15 to 3,022 in 2024-25, a +1.7% annual long-term CAGR. Both CAGRs smooth over a series that was flat through most of the window: annual DC completions held in a 2,417-2,861 band from 2014-15 through 2023-24, and roughly a third of the entire ten-year increase occurred in the final year alone (2023-24 to 2024-25: 2,666 to 3,022, a modest one-year increase of approximately 13%); the smoothed annual rates above should not be read as a steady year-over-year climb. Exact DPT graduations increased +1.8% annually short-term and +2.9% annually long-term. The original 2019-20 sensitivity produced +3.2% annual growth for chiropractic and +1.6% for DPT.
The chiropractic result is a direct counterexample to a simple interpretation that unfavorable early-career burden necessarily produces immediate or sustained contraction in graduations. It suggests that program demand may respond to additional factors such as institutional strategy, profession-specific expectations, geographic distribution, new-program entry, marketing, licensing pathways, or delayed recognition of graduate economics.
Of 18 current chiropractic institutions, 10 had unsuppressed Scorecard professional-practice debt and four-year earnings. The median institution burden ratio was approximately 3.03x, with a range of 2.30x to 4.34x. Across these 10 institutions, short-term burden and graduation growth were positively but nonsignificantly correlated (Spearman rho=+0.41; p=0.244). The long-term association was also positive and reached nominal statistical significance (rho=+0.70; p=0.025). Because the burden estimates are current, the graduation windows are historical and overlapping, the sample is only 10 institutions, and multiple exploratory horizons were examined, the long-term correlation should not be interpreted causally. Its main implication is that chiropractic institutions do not reproduce the pooled negative burden-growth pattern.
The contextual sources explain why a single earnings dataset is insufficient for chiropractic. BLS occupational wage estimates exclude self-employed workers, while College Scorecard tax-linked earnings include positive self-employment income. BLS separately estimates self-employment prevalence and identifies chiropractic as the highest-self-employment occupation among doctoral/professional occupations with numeric estimates. IRS sole-proprietorship data then characterize the business economics of one ownership segment rather than graduate earnings. PSEO offers explicit year-1, year-5, and year-10 earnings trajectories but had limited chiropractic institution coverage in the supplied file. These sources were therefore used to bound interpretation rather than averaged into a composite earnings estimate. Figure 2 displays the broader BLS occupational wage distribution and its substantial within-profession variability.
BLS Employment Projections identify chiropractic as a workforce-structure outlier in addition to its high early-career financial burden. Chiropractors have a 36.3% self-employed share, the highest reported value among the 50 doctoral/professional occupations with numeric self-employment estimates in BLS Table 1.2. The median across those 50 occupations is 3.7%; the next-highest reported values are 12.6% for orthodontists and prosthodontists, 12.4% for general dentists, and 11.3% for lawyers. Physical therapists are 4.2% self-employed, making the chiropractic share approximately 8.6 times higher. Figure 4 highlights this workforce-structure contrast across selected health professions.
| Profession | Typical entry credential | BLS self-employed | BLS median wage | IRS sole-prop returns* |
|---|---|---|---|---|
| Chiropractic | Doctoral/professional | 36.3% | $79,200 | 21,935 |
| Physical therapy | Doctoral/professional | 4.2% | $102,760 | — |
| Dentistry | Doctoral/professional | 12.2% | $170,950 | 56,157 |
| Optometry | Doctoral/professional | 5.0% | $136,570 | 15,382 |
| Pharmacy | Doctoral/professional | 0.3% | $140,910 | — |
| Podiatry | Doctoral/professional | 1.9% | $160,300 | 4,540 |
| Veterinary medicine | Doctoral/professional | 7.2% | $130,100 | — |
| Law | Doctoral/professional | 11.3% | $159,670 | 316,370 |
*IRS: 2023 Schedule C industry returns, not workers; dashes indicate no separately identifiable profession-specific category.
IRS SOI independently identifies 21,935 chiropractic sole-proprietorship returns in tax year 2023. Those returns reported approximately $166,153 in business receipts per return and $55,500 in net income less deficit per return; 77.5% reported positive net income, averaging approximately $74,425 among profitable returns. These are business-return statistics, not graduate earnings or counts of self-employed chiropractors. They are used to characterize the economics of the sole-proprietor segment and to reinforce why wage-only measures incompletely describe a profession with substantial practice ownership.
Chiropractic presents a striking apparent paradox: debt is high, reported early-career earnings are modest, Debt/Y4 remains above 3x, and BBRR4 suggests weak debt resolution, yet national graduations increased +2.5% annually from 2018-19 to 2024-25 and +1.7% annually from 2014-15 to 2024-25. This pattern does not invalidate the pooled burden/graduation association; it demonstrates that educational demand can remain strong even when graduate economics are unfavorable and that profession-specific institutional or behavioral factors matter.
One plausible interpretation is that broad access to federal graduate/professional borrowing reduces near-term price sensitivity: students can finance a high-cost credential before the eventual debt/earnings and repayment consequences are fully observable, allowing graduation volume to remain stable or grow despite unfavorable graduate economics. The current data cannot test that mechanism. They contain no exogenous measure of loan availability, no borrower-level link from financing decisions to subsequent program demand, and no design that identifies a causal effect of federal credit. Other explanations include nonfinancial career preferences, expectations that differ from realized earnings, institutional expansion, geographic demand, licensing structure, and delayed labor-market feedback.
The appropriate interpretation is not that chiropractic has reached a demonstrated tipping point. Rather, the current analysis establishes a pre-policy baseline. If financing becomes materially more restrictive, subsequent changes in applications, enrollment, graduations, tuition, debt, and repayment could test whether credit availability has supported demand despite weak early-career economics.
Primary metric is not lifetime ROI. Debt divided by four-year earnings is an early-career financing-burden indicator. It does not include lifetime earnings, opportunity cost, taxes, discounting, business equity, or a full net-present-value calculation.
Debt is not total educational cost. The Scorecard evaluated-institution debt measure captures Stafford plus Graduate PLUS debt at the evaluated institution and same academic level. For graduate/professional credentials it does not add undergraduate debt, and it excludes accrued interest.
Tuition and debt were deliberately not added. Published tuition is a price measure and debt is a financing measure; adding them would double-count financed tuition. Cost of attendance, grants, and net price are not consistently field-specific across all credential levels.
Scorecard field resolution is four-digit CIP. This is particularly important for physical therapy. CIP 51.23 is PT-dominant but includes occupational therapy and other rehabilitation programs.
Scorecard populations and cohorts differ by measure. The Most Recent field file combines Y1-Y5 earnings metrics from different reporting vintages, cohorts, eligibility rules, worker populations, and inflation reference years. Y2/Y3 are legacy metrics, and the three-year measure included nonworkers. Credential-level earnings are worker-count-weighted averages of institution-program medians rather than individual-level mean earnings. The study therefore presents the five observations as a descriptive earnings series, not as one longitudinal cohort or a calculated within-cohort earnings growth rate.
Graduation analyses are observational. The burden/graduation associations do not establish that financial burden causes changes in graduations. Current Scorecard burden measures overlap with the 2014-15 through 2024-25 graduation periods, so temporal ordering is not causal. Program supply, institutional expansion or closure, labor-market expectations, geography, licensing, accreditation, marketing, and delayed behavioral response may confound the relationship.
Model explanatory power remains low. R² was 0.049 for the short-term model and 0.063 for the long-term model, indicating that most variation in graduation growth remains unexplained by the included variables.
Chiropractic institution analysis is small and exploratory. Only 10 of 18 current institutions had unsuppressed Scorecard debt and four-year earnings. The positive long-term institution correlation (rho=+0.70; p=0.025) was one of several exploratory horizon tests and should not be interpreted as evidence that higher burden causes growth.
PSEO and BLS have self-employment limitations. PSEO primarily reflects unemployment-insurance-covered employment and BLS OEWS excludes self-employed workers, limiting their use for a profession with substantial practice ownership. They were not substituted for Scorecard graduate earnings.
BLS self-employment and IRS SOI are different constructs. BLS estimates worker employment status; IRS counts Schedule C sole-proprietorship returns and excludes other ownership structures such as S corporations and partnerships from this table. IRS business receipts/net income also mix career stages and should not be interpreted as graduate earnings.
BBRR is interval-censored and COVID-sensitive. Many field-of-study repayment values are privacy-suppressed or coarsened into ranges. Midpoints are study sensitivity estimates, not official Scorecard point estimates. Because BBRR categories are hierarchical, borrowers in forbearance are not evaluated for not-making-progress; pandemic-era administrative forbearance can therefore obscure underlying amortization behavior.
The financing-access / tipping-point interpretation is a hypothesis, not a finding. Rising chiropractic graduations can coexist with unfavorable graduate economics for multiple reasons. The study does not identify federal loan availability as the cause and does not establish that a demand inflection is imminent.
NPSAS was not usable for DC-specific estimates. NPSAS:20 PowerStats suppressed chiropractic financial-aid/debt cells because fewer than 30 valid sample cases supported the estimates. The source was de-scoped rather than broadened to an inexact health-professions proxy.
Arete Healthcare operates an MSO/CIN and research program serving independent outpatient practices, including chiropractic and physical therapy practices. This creates a commercial and professional interest in the subject matter. The analysis was designed to preserve separation between reported findings and any institutional interpretation.
No external sponsor funded this analysis. The study used publicly available federal datasets and files downloaded for this project. No patient-level protected health information or proprietary payer data were used.
ChatGPT (OpenAI, GPT-5.6 Sol) materially assisted on September 7-9, 2026 with data transformation, code development, statistical analysis, drafting, figure interpretation and caption development, and document production. The four final In Pictures visualizations were generated by the Author/Analyst in Power BI; AI was used to support interpretation, caption drafting, and placement, not to generate the final visualizations. Claude (Anthropic) subsequently performed independent verification of headline statistics against the analytic workbook and rebuilt this document's structure fresh through the Arete Original Research pipeline (Arete Fresh-Rebuild Process v1.0); Arete's Take (above) is the Author's own writing, not AI-originated, per Governance Framework §12. The Author/Analyst remains responsible for the accuracy and interpretation of the work.
Approved for Release. Figures and reported statistics were independently re-verified against the retained reproducibility workbook (Arete_Education_CBA_Analytic_v1_4.xlsx); no discrepancies were found. This paper was reviewed and approved for publication by the Data Analytics Working Group (DAWG), per the Arete Research Library Governance Framework §7. Human citation/source confirmation was completed prior to publication.
Arete Healthcare. The Education ROI: Cost, Debt, Earnings, and Graduation Trends Across Postsecondary Credentials. Arete Original Research ARETE-OR-2026-010, Version 1.11. Published September 18, 2026. https://doi.org/10.13140/RG.2.2.31764.05767
1. U.S. Department of Education, College Scorecard. Field of Study Data Documentation. https://collegescorecard.ed.gov/files/FieldOfStudyDataDocumentation.pdf
2. National Center for Education Statistics. IPEDS Complete Data Files. https://nces.ed.gov/ipeds/datacenter/DataFiles.aspx
3. National Center for Education Statistics. IPEDS Cost I 2024-25 survey materials. https://nces.ed.gov/ipeds/use-the-data/download-survey-material/2024/cost%20i/package_11_72.pdf
4. U.S. Bureau of Labor Statistics. Occupational Employment and Wage Statistics, May 2025. https://www.bls.gov/oes/tables.htm
5. U.S. Census Bureau. Postsecondary Employment Outcomes documentation. https://lehd.ces.census.gov/data/pseo_documentation.html
6. Internal Revenue Service, Statistics of Income. Nonfarm Sole Proprietorship Statistics. https://www.irs.gov/statistics/soi-tax-stats-nonfarm-sole-proprietorship-statistics
7. National Center for Education Statistics. NPSAS:20 Data File Documentation / PowerStats. https://nces.ed.gov/use-work/resource-library/report/users-manualdata-file-documentation/2019-20-national-postsecondary-student-aid-study-npsas20-data-file-documentation
The terms below are used throughout this paper; each is also introduced in context in Methods.
| Term | Definition |
|---|---|
| CAGR | Compound annual growth rate. Primary short-term CAGR links 2018-19 to 2024-25: (C2024-25 / C2018-19)^(1/6) - 1. Primary long-term CAGR links 2014-15 to 2024-25: (C2024-25 / C2014-15)^(1/10) - 1. The 2019-20 to 2024-25 five-year CAGR is retained as a COVID-overlap sensitivity. |
| CBA | Cost-benefit analysis. The current paper uses an early-career debt/earnings burden proxy; it does not claim a complete lifetime cost-benefit or net-present-value estimate. |
| CIP | Classification of Instructional Programs, the NCES taxonomy for fields of study. Scorecard field outcomes are generally four-digit CIP; IPEDS graduations provide six-digit detail. |
| Debt / 4-year earnings | Mean cumulative evaluated-institution debt divided by national median earnings four years after graduation. Lower values are more favorable. |
| Evaluated-institution debt | College Scorecard Stafford plus Graduate PLUS debt originated at the evaluated institution for the same academic level; it is not total lifetime educational debt. |
| HC3 | A heteroskedasticity-consistent standard-error estimator that is robust to nonconstant residual variance and gives additional protection against leverage in finite samples. |
| IPEDS | Integrated Postsecondary Education Data System, NCES institutional census data used here for graduations, program classification, and tuition/cost measures. |
| NPSAS | National Postsecondary Student Aid Study. NPSAS:20 was evaluated but DC-specific estimates were suppressed because fewer than 30 valid sample cases supported them. |
| OEWS | BLS Occupational Employment and Wage Statistics. OEWS wage estimates exclude self-employed workers. |
| PSEO | Census Postsecondary Employment Outcomes, used as an earnings sensitivity source where participating institutions are represented. |
| Self-employed share | BLS Employment Projections estimate of the percentage of workers in an occupation who are self-employed. This is the study's prevalence measure for workforce structure. |
| SOI | IRS Statistics of Income. The 2023 nonfarm sole-proprietorship tables summarize Schedule C business returns by industry. |
| Sole-proprietorship return | An IRS Schedule C business tax return. It is not equivalent to one graduate or one worker and cannot be divided by BLS employment to estimate sole-proprietor prevalence. |
| Professional-practice degree | IPEDS/Scorecard credential category for degrees preparing graduates for professional practice; distinguished from doctoral research/scholarship and other doctorates. |
| Field-level burden ratio | For a field/credential, pooled mean evaluated-institution debt divided by the field’s national median earnings four years after graduation. This is the primary field-level burden measure used in cross-field analyses. |
| Program-level burden ratio | For an institution-program observation, institution-specific evaluated-institution mean debt divided by institution-specific median earnings four years after graduation. Within-field program P25/median/P75 values use only unsuppressed numeric observations and are not interchangeable with the field-level pooled ratio. |
| Scorecard earnings series | Reported post-graduation earnings observations at years 1 through 5. Y2 and Y3 are legacy metrics from earlier Scorecard vintages; Y3 used different eligibility rules and included nonworkers. Field summaries use weighted program medians for Y1/Y2/Y3/Y5 and the national field median for Y4. Credential summaries use worker-count-weighted averages of institution-program medians for Y1-Y5. The five observations are descriptive and are not a same-cohort longitudinal series. |
| Worker-count-weighted program-median earnings | A constructed aggregate equal to sum(institution-program median earnings x the corresponding Scorecard worker count) divided by sum(worker counts). For credential-level summaries, this gives larger reporting populations proportional influence and avoids treating every field as equally sized. It is not an individual-level mean earnings estimate. |
| BBRR / BBRR4 | Borrower-Based Repayment Rate; BBRR4 classifies federal borrowers four years after entering repayment into a hierarchical, mutually exclusive status. |
| Not making progress | BBRR status for borrowers making regular payments whose total outstanding balance still exceeds original balances, after excluding default, delinquency, forbearance, and deferment. |
| Positive amortization | Study-defined descriptive construct equal to BBRR making progress + paid in full; not an official Scorecard metric. |
| BBRR midpoint | Study sensitivity estimate for a privacy-coarsened Scorecard interval; not an official Scorecard point estimate. |
| Reported burden compression | Debt/Y1 minus Debt/Y4. Descriptive cross-cohort measure of how much the reported debt/earnings ratio differs between the Y1 and Y4 earnings observations; not within-person loan amortization. |
| Graduation-financing paradox | Coexistence of unfavorable graduate economics with stable or growing graduations. The term describes the observation; it does not establish that federal financing caused the growth or that a tipping point is imminent. |
| Field | 4-year earnings | Debt | Debt/earnings | Graduation CAGR: short / long |
|---|---|---|---|---|
| Alternative/complementary medicine | $41,692 | $143,279 | 3.44x | -4.6% / +3.8% |
| Chiropractic | $61,163 | $196,057 | 3.21x | +2.5% / +1.7% |
| Medicine | $112,460 | $183,753 | 1.63x | +7.7% / +5.2% |
| Dentistry | $170,812 | $264,439 | 1.55x | +1.7% / +1.8% |
| Optometry | $131,525 | $182,745 | 1.39x | -0.5% / +0.8% |
| Veterinary medicine | $131,316 | $178,661 | 1.36x | — / — |
| Rehabilitation/therapeutic professions* | $81,641 | $109,155 | 1.34x | +4.8% / +5.2% |
| Law | $107,773 | $121,285 | 1.13x | +1.0% / -1.0% |
| Communication disorders | $85,242 | $90,919 | 1.07x | +0.7% / +2.3% |
| Clinical/counseling psychology | $70,514 | $74,819 | 1.06x | +4.5% / +1.6% |
| Pharmacy | $138,348 | $141,199 | 1.02x | -4.8% / -2.6% |
| Theological/ministerial studies | $63,395 | $54,231 | 0.86x | -1.0% / -0.8% |
| Legal professions/studies, other | $86,371 | $68,397 | 0.79x | -2.8% / -4.0% |
| Nursing | $133,099 | $77,402 | 0.58x | +10.4% / +13.5% |
| Educational administration/supervision | $78,026 | $37,322 | 0.48x | +12.0% / +0.3% |
| Allied health Dx/intervention | $127,648 | $54,195 | 0.42x | +10.6% / — |
*Four-digit CIP 51.23; PT-dominant but mixed. Short-term CAGR uses 2018-19 to 2024-25; long-term CAGR uses 2014-15 to 2024-25. CAGR is shown only when the relevant baseline and 2024-25 graduations are each at least 30; em dash indicates insufficient baseline volume.
| Field | Y1 | Y2* | Y3* | Y4 | Y5 |
|---|---|---|---|---|---|
| Alternative/complementary medicine | $32,880 | $22,877 | $31,507 | $41,692 | $43,856 |
| Chiropractic | $49,051 | $39,382 | $43,714 | $61,163 | $60,133 |
| Medicine | $65,207 | $57,606 | $67,781 | $112,460 | $172,595 |
| Dentistry | $116,891 | $108,540 | $129,383 | $170,812 | $156,920 |
| Optometry | $98,468 | $94,562 | $107,537 | $131,525 | $118,840 |
| Veterinary medicine | $100,548 | $78,521 | — | $131,316 | $120,712 |
| Rehabilitation/therapeutic professions* | $74,773 | $70,058 | $73,183 | $81,641 | $82,754 |
| Law | $86,981 | $73,858 | $88,820 | $107,773 | $120,998 |
| Communication disorders | $70,330 | $62,228 | $63,467 | $85,242 | $70,624 |
| Clinical/counseling/applied psychology | $72,150 | $60,459 | $70,500 | $70,514 | $87,331 |
| Pharmacy | $101,042 | $117,728 | $124,534 | $138,348 | $142,528 |
| Theological/ministerial studies | $57,750 | $44,220 | $47,490 | $63,395 | $58,824 |
| Legal professions/studies, other | $104,329 | $64,438 | $55,787 | $86,371 | $84,844 |
| Registered nursing | $123,936 | $117,333 | $122,486 | $133,099 | $140,629 |
| Educational administration/supervision | $69,286 | $65,324 | $70,755 | $78,026 | $83,047 |
| Allied health diagnostic/intervention | $105,241 | $100,993 | $104,235 | $127,648 | $121,762 |
*Y2 and Y3 are legacy Scorecard metrics from earlier reporting vintages; Y3 used different eligibility rules and included nonworkers. Y1/Y2/Y3/Y5 are weighted averages of institution-program medians; Y4 is the national field median. The rows are ordered by debt-to-four-year-earnings burden in the preceding table, not by earnings. These observations are not a same-cohort longitudinal series. CIP 51.23 is PT-dominant but mixed.
| Field | Program records | Programs with burden | Field burden | Program P25 | Program median | Program P75 |
|---|---|---|---|---|---|---|
| Alt./comp. medicine | 28 | 5 | 3.44x | 2.75x | 3.46x | 3.78x |
| Chiropractic | 17 | 10 | 3.21x | 2.89x | 3.03x | 3.49x |
| Medicine | 213 | 137 | 1.63x | 1.37x | 1.78x | 2.04x |
| Dentistry | 71 | 53 | 1.55x | 1.19x | 1.42x | 1.67x |
| Optometry | 23 | 16 | 1.39x | 1.16x | 1.26x | 1.41x |
| Veterinary medicine | 46 | 30 | 1.36x | 1.21x | 1.37x | 1.68x |
| Rehab./therapeutic professions* | 278 | 52 | 1.34x | 1.01x | 1.34x | 1.60x |
| Pharmacy | 143 | 82 | 1.02x | 0.77x | 1.00x | 1.30x |
Field burden is the field-level pooled evaluated-institution debt divided by national median four-year earnings. Program P25/median/P75 values use institution-specific debt divided by institution-specific four-year earnings among unsuppressed records; the analytic unit therefore differs from the field-level burden ratio. *CIP 51.23 is PT-dominant but mixed. Small program-burden N, particularly alternative/complementary medicine (N=5), should be interpreted cautiously.
| Field | BBRR4 N | FBR midpoint (coverage) | NOPROG midpoint (coverage) | +amort. midpoint (coverage) | Debt/Y1 → Debt/Y4 | Short / long CAGR |
|---|---|---|---|---|---|---|
| Alt./comp. medicine | 90 | 90% (100%) | — (—) | — (—) | 4.36x → 3.44x | -4.6% / +3.8% |
| Chiropractic | 2,272 | 74% (95%) | 17% (87%) | 13% (76%) | 4.00x → 3.21x | +2.5% / +1.7% |
| Medicine | 21,940 | 47% (97%) | 26% (94%) | 26% (83%) | 2.82x → 1.63x | +7.7% / +5.2% |
| Dentistry | 6,500 | 37% (99%) | 18% (94%) | 44% (91%) | 2.26x → 1.55x | +1.7% / +1.8% |
| Optometry | 1,168 | 36% (100%) | 18% (83%) | 44% (72%) | 1.86x → 1.39x | -0.5% / +0.8% |
| PT-dominant rehab* | 3,794 | 44% (92%) | 20% (79%) | 35% (73%) | 1.46x → 1.34x | +4.8% / +5.2% |
| Law | 39,131 | 46% (98%) | 23% (97%) | 30% (94%) | 1.39x → 1.13x | +1.0% / -1.0% |
| Pharmacy | 14,866 | 36% (100%) | 17% (97%) | 45% (94%) | 1.40x → 1.02x | -4.8% / -2.6% |
Midpoints are analytic sensitivity estimates of Scorecard coarsened ranges and are weighted by BBRR4 borrower N; status-specific coverage appears in parentheses. Because categories are hierarchical, values should not be summed across rows without accounting for differing coverage. Forbearance is COVID-sensitive.
| Institution | 4-year earnings | Debt | Debt/earnings | Graduation CAGR: short / long |
|---|---|---|---|---|
| Institution 08 | $73,854 | $169,762 | 2.30x | -4.3% / -3.6% |
| Institution 09 | $69,126 | $187,189 | 2.71x | -3.0% / -0.3% |
| Institution 06 | $62,214 | $178,039 | 2.86x | -1.3% / -0.4% |
| Institution 04 | $60,806 | $179,895 | 2.96x | +7.9% / +2.8% |
| Institution 01 | $60,506 | $180,918 | 2.99x | +2.0% / -1.3% |
| Institution 05 | $62,795 | $192,822 | 3.07x | +9.5% / +7.1% |
| Institution 07 | $56,605 | $182,187 | 3.22x | +16.1% / +3.7% |
| Institution 03 | $55,900 | $199,949 | 3.58x | -3.1% / +5.4% |
| Institution 10 | $57,734 | $241,902 | 4.19x | +1.2% / +1.9% |
| Institution 02 | $44,874 | $194,817 | 4.34x | +2.1% / +7.0% |
Debt values shown are institution-level Scorecard evaluated-institution means. Graduation CAGR is shown as short-term 2018-19 to 2024-25 / long-term 2014-15 to 2024-25. Institution-level results are exploratory and subject to Scorecard suppression, program changes, and overlap between current burden measurement and historical graduation periods. Institution labels are randomly assigned and consistent across the institution-level Appendix tables. Because the underlying federal data are public and the number of chiropractic programs is small, masking reduces emphasis on individual schools but does not guarantee anonymity.
| Institution | BBRR4 N | Forbearance | Not making progress | Making progress | Paid in full | Short / long CAGR |
|---|---|---|---|---|---|---|
| Institution 06 | 662 | >=0.80 | <=0.20 | <=0.20 | <=0.20 | -1.3% / -0.4% |
| Institution 10 | 552 | 0.65 - 0.69 | 0.25 - 0.29 | <=0.05 | <=0.05 | +1.2% / +1.9% |
| Institution 04 | 332 | 0.60 - 0.69 | 0.20 - 0.29 | <=0.10 | <=0.10 | +7.9% / +2.8% |
| Institution 08 | 243 | 0.60 - 0.79 | <=0.20 | PS | <=0.20 | -4.3% / -3.6% |
| Institution 07 | 197 | 0.60 - 0.79 | PS | PS | PS | +16.1% / +3.7% |
| Institution 09 | 181 | 0.60 - 0.79 | <=0.20 | <=0.20 | <=0.20 | -3.0% / -0.3% |
| Institution 03 | 105 | PS | PS | PS | PS | -3.1% / +5.4% |
Source values are shown as published to preserve privacy-coarsened ranges. PS = privacy suppressed. Only seven chiropractic programs have numeric BBRR4 cohort N in the supplied Scorecard file; category visibility is more limited. Institution labels use the same randomized identifiers as the preceding table.
| Rank | Occupation | Self-employed | Median wage | Employment (000s) |
|---|---|---|---|---|
| 1 | Chiropractors | 36.3% | $79,200 | 61.7 |
| 2 | Orthodontists | 12.6% | $289,140 | 7.1 |
| 3 | Prosthodontists | 12.6% | $311,180 | 1.0 |
| 4 | Dentists, general | 12.4% | $170,950 | 141.9 |
| 5 | Oral and maxillofacial surgeons | 12.3% | $352,220 | 5.6 |
| 6 | Lawyers | 11.3% | $159,670 | 863.7 |
| 7 | Veterinarians | 7.2% | $130,100 | 91.1 |
| 8 | Physicists | 7.0% | $172,250 | 23.2 |
| 9 | Dentists, all other specialists | 7.0% | $224,990 | 5.8 |
| 10 | Astronomers | 5.8% | $128,820 | 2.4 |
BLS Table 1.2 includes 73 detailed occupations for which the typical entry education is a doctoral or professional degree; 50 have numeric self-employment estimates. Ranking is among those 50 numeric observations.
| Profession | Credential context | Self-employed | BLS median wage | IRS sole-prop returns | IRS net income / return |
|---|---|---|---|---|---|
| Chiropractic | Doctoral/professional | 36.3% | $79,200 | 21,935 | $55,500 |
| Physical therapy | Doctoral/professional | 4.2% | $102,760 | — | — |
| Medicine | Doctoral/professional | 3.7% | — | 181,294 | $70,959 |
| Dentistry | Doctoral/professional | 12.2% | $170,950 | 56,157 | $86,566 |
| Optometry | Doctoral/professional | 5.0% | $136,570 | 15,382 | $44,367 |
| Pharmacy | Doctoral/professional | 0.3% | $140,910 | — | — |
| Podiatry | Doctoral/professional | 1.9% | $160,300 | 4,540 | $90,410 |
| Veterinary medicine | Doctoral/professional | 7.2% | $130,100 | — | — |
| Clinical/counseling psychology | Doctoral/professional | 4.8% | $100,580 | — | — |
| Acupuncture | Master's | 46.2% | $76,040 | — | — |
| Law | Doctoral/professional | 11.3% | $159,670 | 316,370 | $59,807 |
| Architecture | Bachelor's | 14.0% | $99,280 | 139,796 | $10,993 |
| Accounting / CPA context | Bachelor's | 5.5% | $83,680 | 28,476 | $65,773 |
BLS self-employment is an occupational worker-status measure. IRS SOI figures are 2023 Schedule C business returns by industry and are shown only where a reasonably specific industry mapping is available. IRS net income is business net income less deficit per return, not graduate or practitioner earnings.
IRS Statistics of Income Schedule C data provide a practice-economics view that complements graduate earnings measures. The table below reports 2023 aggregate income-statement items for Offices of Chiropractors. Amounts are IRS-reported thousands of dollars; per-return values use the 21,935 chiropractic sole-proprietorship returns as a common denominator for comparability, and percentages are relative to total business receipts. These are business-return economics, not chiropractor personal earnings, and they exclude incorporated practices, S corporations, partnerships, and other non-Schedule C ownership structures. The positive-net-income subset contained 16,999 returns; average net income among those profitable returns was approximately $74,425. Source: IRS Statistics of Income, Nonfarm Sole Proprietorship Statistics, Tax Year 2023.
| Line item | Amount ($000) | Per all returns ($) | % of receipts |
|---|---|---|---|
| Number of returns | 21,935 | — | — |
| Business receipts, total | 3,644,566 | $166,153 | 100.0% |
| Income from sales and operations | 3,579,598 | $163,191 | 98.2% |
| Other business income (loss) | 64,968 | $2,962 | 1.8% |
| Business deductions, total | 2,427,169 | $110,653 | 66.6% |
| Cost of sales and operations, total | 187,883 | $8,565 | 5.2% |
| Advertising expenses | 97,919 | $4,464 | 2.7% |
| Car and truck expenses | 53,726 | $2,449 | 1.5% |
| Commissions | 11,836 | $540 | 0.3% |
| Contract labor | 129,400 | $5,899 | 3.6% |
| Depreciation | 77,841 | $3,549 | 2.1% |
| Employee benefit programs | 16,250 | $741 | 0.4% |
| Insurance | 64,559 | $2,943 | 1.8% |
| Legal and professional services | 66,781 | $3,044 | 1.8% |
| Meals deducted | 20,075 | $915 | 0.6% |
| Mortgage interest | 267 | $12 | 0.0% |
| Other interest paid on business indebtedness | 21,567 | $983 | 0.6% |
| Office expenses | 125,298 | $5,712 | 3.4% |
| Pension and profit-sharing plans | 5,762 | $263 | 0.2% |
| Rent paid on machinery and equipment | 18,788 | $857 | 0.5% |
| Rent paid on other business property | 313,301 | $14,283 | 8.6% |
| Repairs | 32,034 | $1,460 | 0.9% |
| Supplies | 182,796 | $8,334 | 5.0% |
| Salaries and wages | 469,477 | $21,403 | 12.9% |
| Taxes paid | 51,585 | $2,352 | 1.4% |
| Travel | 30,643 | $1,397 | 0.8% |
| Utilities | 95,347 | $4,347 | 2.6% |
| Other business expenses | 341,013 | $15,547 | 9.4% |
| Home office business deductions, total | 13,022 | $594 | 0.4% |
| Net income less deficit | 1,217,397 | $55,500 | 33.4% |
| Net income | 1,265,159 | $57,678 | 34.7% |
| Deficit | 47,763 | $2,177 | 1.3% |
| Qualified business deduction | 138,297 | $6,305 | 3.8% |
Note: “Net income” is an aggregate positive-income amount reported by IRS; the per-all-returns column intentionally retains the common 21,935-return denominator for comparability. Net income less deficit is the more appropriate aggregate business-income measure across all returns. Qualified business deduction is a tax deduction, not a business operating expense.