What Is Degree Attainment? (Census Metrics)
Degree attainment is the share of a defined population that has completed a stated level of education. In U.S. Census comparisons, it often means the share of people age 25 and older with a bachelor’s degree or higher. It is not an enrollment rate, a college graduation rate, or a count of degrees awarded. The population, degree threshold, place, source, and year all matter.
If you see a claim such as “40% of adults have a degree,” the number is only useful if you know which adults, which degree, and where the data came from. A rate for adults with a bachelor’s degree or higher answers a different question from a rate that includes associate degrees or certificates.
I look first for the measure’s definition, not just its headline percentage. That small step helps prevent misleading comparisons when you are reading a report about a city, state, or college-going population.
What Census degree-attainment rates measure
Degree attainment is the share of a stated population that has completed a specified degree level. In a common U.S. Census Bureau measure, the population is people age 25 and older, and the threshold is a bachelor’s degree or higher. The result describes people in a place, not the performance of one college.
The American Community Survey (ACS) table B15003, called “Educational Attainment,” reports education levels for the population age 25 and older. Its total is the denominator for calculations using this table. The ACS does not set one universal “degree-attainment rate”: the result changes with the degree threshold, geography, data product, and reference period you choose.
A Census rate is a population measure. It tells you what share of residents in the defined group report having completed a given level of education. It does not tell you how many students are currently in college or how many degrees colleges awarded that year.
For example, a state rate for bachelor’s degree or higher could include adults who earned their degrees elsewhere and moved to the state. It also includes people who completed degrees many years ago. You cannot use that figure alone to judge a particular university or program.
What the age-25-and-older universe means
The universe is the population covered by a table. B15003’s universe is people age 25 and older. This age boundary gives people time to complete education, but it does not mean every person in the group had the same chance to attend college or followed the same path.
Use the universe stated in the table, and include it when you report a rate. If another source measures people age 18 and older, or only enrolled students, its percentage is not directly comparable to B15003, even when the label sounds similar.
Degree categories and how to calculate a rate
A degree threshold states which education categories count in the numerator. To calculate a rate, add the table categories that meet that threshold, divide by the table’s total population, and multiply by 100. State the threshold, age group, geography, and ACS period so another reader can understand the result.
In B15003, the categories for a bachelor’s degree or higher are bachelor’s, master’s, professional-school, and doctorate degrees. The table’s detailed estimates are numbered: bachelor’s is B15003_022, master’s B15003_023, professional-school B15003_024, and doctorate B15003_025. The total population is B15003_001.
Formula:
Bachelor’s degree or higher rate = (bachelor’s + master’s + professional-school + doctorate estimates) ÷ total population × 100
For a measure that includes associate degrees, add that category too. Be precise in the name: “associate degree or higher” and “bachelor’s degree or higher” are different measures. If you say “any degree,” explain whether an associate degree counts.
| Measure | Categories included in the numerator | What the result means |
|---|---|---|
| Bachelor’s degree or higher | Bachelor’s, master’s, professional-school, doctorate | Share of the stated population with at least a bachelor’s degree |
| Associate degree or higher | Associate, bachelor’s, master’s, professional-school, doctorate | Share with at least an associate degree |
| Some college, no degree | A separate educational-attainment category | Share who attended college but did not report a completed degree |
| Current enrollment | Not calculated from B15003 attainment categories | Share enrolled, if measured by a suitable enrollment source |
A certificate is not the same as a degree in this calculation. Neither “some college” nor current enrollment belongs in a degree-completion numerator. Keep those categories separate, even when a report groups them under a broader label such as postsecondary education.
A worked calculation using hypothetical numbers
Suppose a hypothetical county’s B15003 total is 100,000 people age 25 and older. Its four degree estimates are 24,000 bachelor’s, 12,000 master’s, 2,000 professional-school, and 2,000 doctorate degrees.
Add the four estimates: 24,000 + 12,000 + 2,000 + 2,000 = 40,000. Divide 40,000 by 100,000 and multiply by 100. The result is 40% with a bachelor’s degree or higher. These numbers are for illustration, not a reported county result.
Choose the right ACS source and compare estimates carefully
The ACS provides estimates for different release periods and geographic areas. A one-year estimate gives a shorter reference period and is generally available for areas with populations of 65,000 or more. Five-year estimates combine data across five years and cover smaller areas. Record the ACS product and period with every rate.
Use the ACS table and release that fit your question. For a detailed breakdown, start with B15003 on data.census.gov. Check the table title, universe, geography, and release period before copying figures. The Census Bureau’s ACS guidance explains data products and their use.
| Data choice | Typical use | Key caution |
|---|---|---|
| ACS 1-year estimate | Looking at places with populations of 65,000 or more, with a shorter data period | Smaller places may not have a 1-year estimate |
| ACS 5-year estimate | Looking at smaller areas or seeking data for a broader range of geographies | The estimate represents a five-year period, not one single year |
| Overlapping ACS 5-year releases | Reviewing estimates across successive releases | Releases share years, so they are not independent annual observations |
A five-year estimate smooths data across its period. For example, two releases with overlapping years share some of the same underlying survey years. Do not describe their difference as a clean year-to-year change or treat each estimate as fully independent.
Check uncertainty as well as the percentage
The ACS uses a sample, so its estimates have uncertainty. Many ACS tables provide a margin of error (MOE) with an estimate. When a difference between two places or periods is small, check the MOEs before treating it as meaningful. A percentage without its source and uncertainty can suggest more precision than the data support.
When calculating a rate from several categories, do not simply add their MOEs and present the result as the rate’s MOE. Follow Census Bureau guidance for combining estimates and margins of error, or use a published rate that matches your question. Keep the estimate type and method in your notes.
Worked scenarios: interpret the number without overclaiming
A worked scenario shows how the measure behaves when the population, threshold, or source changes. These examples are hypothetical, not findings about real places. They demonstrate why two percentages can both be correct yet answer different questions.
Scenario 1: Two degree thresholds. Imagine a report says 45% of adults age 25 and older have an associate degree or higher, while another says 32% have a bachelor’s degree or higher. The figures can both be valid because the first includes associate degrees. They are not competing estimates of the same threshold.
Scenario 2: A city and a college. A city’s 25-and-older population has a 35% bachelor’s-or-higher rate. That does not mean the city’s local college graduates 35% of its students. The city rate includes residents educated at many institutions and may include people who moved there after graduation.
Scenario 3: One-year and five-year data. A researcher compares a city’s one-year estimate with a county’s five-year estimate and calls one higher. The percentages may describe different geographies and time periods, and the five-year figure is an average across multiple survey years. First align the places, measure, and period as closely as the data allow.
Use a comparison checklist before drawing a conclusion:
- Population: Are both figures for people age 25 and older, or the same stated age group?
- Threshold: Do both count bachelor’s degree or higher, or do they include different credentials?
- Geography: Are both figures for the same kind of place and boundary?
- Source and period: Are both from the ACS, and are their release periods clear?
- Uncertainty: Do the estimates and margins of error support the difference you describe?
If any answer differs, explain the difference instead of presenting the rates as a direct comparison.
What degree attainment can—and cannot—show
Degree attainment can describe the education profile of a population. It can help compare places when the same threshold, age universe, source, and period are used. It does not, on its own, explain why one place has a higher rate or predict an individual student’s outcome.
It is also different from institutional and labor-market measures. IPEDS graduation measures focus on students at institutions under defined reporting rules. Enrollment statistics count people attending school, while degrees awarded count credentials issued during a period. Census attainment instead counts people in a population who report having completed an education level.
Employment outcomes require separate data. A higher attainment rate does not by itself prove that residents earn more, that a specific degree caused an employment result, or that a local college’s programs drove the rate. Those claims need evidence designed to answer those questions.
When you write or share a figure, use a label such as: “Share of county residents age 25 and older with a bachelor’s degree or higher, ACS 5-year estimate, [release period].” This makes the numerator, denominator, source, and period easier to verify.
Conclusion: report the definition with the rate
Degree attainment is a completion measure for a defined population, not a count of students enrolled or degrees awarded. The most common Census comparison uses adults age 25 and older and counts bachelor’s degrees and higher credentials. Always state the geography, threshold, ACS release, and estimate type.
For a reliable comparison, retrieve B15003, add only the categories that fit your threshold, divide by the table total, and report the result as a percentage. Check uncertainty and avoid attributing a population pattern to a college without separate evidence.
Frequently asked questions
These answers clarify common questions about Census degree-attainment metrics. Each one depends on the definition and source used, so check the table universe and data period before comparing a reported percentage.
What does degree attainment mean?
It is the share of a stated population that has completed a specified degree level.
What does Census degree attainment usually measure?
A common measure is the share of people age 25 and older with a bachelor’s degree or higher.
What is the denominator in ACS table B15003?
The denominator is the table’s total population, whose universe is people age 25 and older.
Which B15003 categories count as bachelor’s degree or higher?
Bachelor’s, master’s, professional-school, and doctorate degrees count.
Does “some college” count as degree attainment?
No. “Some college, no degree” is a separate category and does not count as a completed degree.
Do certificates count as degrees in the B15003 calculation?
No. Do not treat certificates as degrees in this calculation.
Is degree attainment the same as college enrollment?
No. Attainment describes completed education; enrollment describes attendance during a stated period.
Is a Census attainment rate a college graduation rate?
No. It describes residents of a place, not the graduation performance of one institution.
When should I use ACS one-year estimates?
They are generally available for areas with populations of 65,000 or more. Check the release and geography before use.
Can I compare consecutive ACS five-year estimates as annual changes?
Not as independent annual observations. Their periods overlap, so they share survey years.
What should I report with an attainment percentage?
State the population, degree threshold, geography, ACS source and period, and estimate type.
(This article was written by one of our staff writers, Nathan Sterling. Visit our Meet the Team page.)
