Results From the Survey of Community College Outcomes Extended Pilot

After three years of developing partnerships with community colleges across the Fifth District, we are proud to release our first-year results from our extended pilot of the Survey of Community College Outcomes.

When the Richmond Fed launched this initiative, the most common question we heard from community colleges was, "Why does the Richmond Fed care about community colleges?" The answer is that the Richmond Fed cares about workforce development, and community colleges play a vital role across the Fifth District providing individuals with the skills needed to participate in the workforce.

The Need for Better Data on Community Colleges

The Integrated Postsecondary Education Data System (IPEDS) has long been the primary source of data on higher education outcomes. IPEDS provides rich information about postsecondary education. However, IPEDS graduation data are not suited to assess community colleges: It uses a first-time, full-time cohort to calculate graduation rates, and students must complete a degree or certificate of more than 16 hours to be considered a "success." This measure of success doesn't fit the reality of many students who attend community colleges. (See "Why We Need a New Outcome Measure in the Community College Space.")

According to IPEDS, Fifth District community colleges graduate about 30 percent of their students, which is less than half of the average four-year institution graduation rate. In conversations, community college leaders often mentioned this issue of mismeasurement of success, which is why state community college systems and the community colleges themselves created their own metrics. However, these metrics are inconsistent across institutions, making it difficult to compare outcomes across schools and states. In order to observe Fifth District community college outcomes across the five states in our district (North Carolina, South Carolina, Maryland, Virginia, and most of West Virginia), we decided to collect consistent data that would enable us to calculate a new success metric: the Richmond Fed Community College Success Rate.

Principles of Our Data Collection Program

Understanding how community colleges prepare their students for the workforce was a big undertaking. When we set out on this journey, we had six guiding principles:

1. Community College-Focused: The metrics that we collect and publish are derived from direct conversations with community colleges.
2. Independent: Although we received significant input from community colleges, our goal was to create a program that was independent and unbiased.
3. Universally Available: We are committed to open and easy public access to the data we collect.
4. Nonburdensome: Community colleges are already providing a lot of information to different reporting agencies. Since this program is voluntary, we reduced response burden by focusing on metrics most relevant for measuring success.
5. Longevity: One of the strengths of IPEDS is that data are updated annually to assess change over time. Our goal is to continue this annual data collection program for the foreseeable future.
6. Good Coverage: The role a community college plays in the workforce pipeline differs by location, industry demands, and student demands. We worked hard to get all the schools in our district to participate to ensure we are covering the experiences of all community colleges.

The Richmond Fed's Community College Metrics

As mentioned earlier, the first task was accurately counting the cohort, or the denominator for our calculations. The Richmond Fed has a more expanded definition than IPEDS in order to include non-first-time and part-time students. Additionally, we expanded the success measurement time frame to include students who are taking longer to complete their learning journey. This is especially important for part-time students.

The Richmond Fed wanted to generate a statistic that encompasses the various ways a credit student can achieve success. This rate can be compared across institutions, similar to graduation rates. A credit student is considered successful if they:

• Receive an award (associate degree, certificate, licensure, industry-recognized credential),
  or
• Transfer to a four-year institution before receiving an award,
  or
• Continue to persist (still enrolled and in good standing but taking longer than four years to complete degree or certificate).

The Richmond Fed's Survey of Community College Outcomes also collects information on more types of students. In addition to the "traditional" credit degree-seeker, we attempt to collect data on non-credit students, workforce training, and high school students.

You can see the full list of metrics we collect by school on our website.

Community College Participation

Through partnerships with the state community college system offices in South Carolina, Virginia, and West Virginia, and the Maryland Community College Research Group, we achieved near universal coverage for community colleges within those four states. We are working with North Carolina on participating in future iterations of this program.

The Richmond Fed Success Rate

The Richmond Fed's student cohorts are larger than IPEDS, which is not surprising considering the Richmond Fed's definition is more inclusive. For example, in Maryland, the cohort was 33,317 more students than what IPEDS reports. This difference was primarily driven by the predominance of part-time students who make up the majority of enrolled students at community colleges but are not included in the IPEDS cohort.

The Richmond Fed Success Rate also reported higher student success compared to the traditional IPEDS graduation rate. The difference was especially noticeable when looking at the bottom end of the range of success rates. While the IPEDS graduation rate was as low as 11 percent at Fifth District community colleges, the lowest Richmond Fed Success Rate was 33 percent.

So, what is driving student success? Interestingly, the story varies by state. South Carolina and West Virginia students were more likely to earn an award at the community college, while Maryland and Virginia students were more likely to transfer to a four-year institution prior to earning an award. In addition, South Carolina has a higher proportion of students who were persisting. (See chart below.) It's important to note that some of these differences were likely driven by state or institution policies. For example, some guaranteed admissions agreements between community colleges and four-year schools require students to complete the associate degree prior to transferring (e.g., the University of South Carolina), while others may only require as many as 30 credits for guaranteed admission (e.g., Virginia Commonwealth University).

Success also varied based on whether students received Pell Grants. Non-Pell students had higher rates of success than Pell students — except South Carolina where the success rates were identical. (See chart below.) This is not uniform across schools, so we plan to do additional research to see what policies or services promote higher levels of success for Pell students.

Dual Enrollment Metrics

High school students across our district are attending community colleges to attain college credits via dual enrollment programs. Dual enrollment has been growing at nearly all Fifth District community colleges, and we regularly hear from presidents that this is the fastest growing student segment. (See "The Role of Community Colleges in Educating High School Students.") In Virginia, nearly 22 percent of all credit students in community colleges were dually enrolled high school students. In Maryland, which has the lowest percentage, almost 14 percent of their credit students were high schoolers. (See chart below.) We expect these dual enrollment percentages to grow in the coming years, in part because Maryland and West Viriginia have recently passed laws that allow high schoolers to attend community colleges tuition free.

Our data show that dual enrollment students were very successful, which we measure as the percentage of attempted credit hours that were successfully attained. In Maryland and Virginia, dually enrolled high schoolers were more likely to successfully attain attempted credits than non-dual enrollment students, with rates over 85 percent in both states. (See chart below.) In West Virginia, dual enrollment students had slightly lower credit attainment rates than their non-dual enrollment counterparts, but students were still remarkably successful in their attempts to earn college credit while still in high school.

Non-Credit Data

Obtaining data for non-credit programs is difficult since very little data on these students exist. (See "The Non-Credit Information Gap.") For many institutions, non-credit program data are stored in different systems than for-credit program data. Additionally, non-credit programs can vary from non-workforce training continuing education to courses that result in high-tech credentials, making it tricky to define what should be counted in our data collection program. Still, we pursued collecting non-credit data with the knowledge that these data would not be as clean as other categories of students.

Improving data collection for non-credit students is critical because they make up a significant portion of overall enrollment, partly because of policy decisions. Maryland is unique because it funds both institutions and students identically regardless of credit or non-credit status. Based on data we collected on non-credit programs that result in industry-recognized credentials, 39 percent of Maryland students were enrolled in these programs. South Carolina has made very large investments into shorter-term, industry-recognized, non-credit programs across the state, and 27 percent of students were enrolled in non-credit programs. (See chart below.)

Community College Enrollment by Gender

One interesting observation in our data is the gender distribution across credit, dual enrollment, and non-credit students. Female students tend to make up a larger portion of the schools' student bodies, especially among credit and dual credit students. However, the distribution of gender in non-credit programs shows a different pattern. In Maryland, non-credit enrollment is dominated by female students, but non-credit in South Carolina, Virginia, and West Virginia is majority male. Why is the gender distribution different for non-credit programs? Do men have a stronger preference for shorter-term credentials? Are male-dominated fields more likely to only require a shorter-term credential? This will be a focus of additional research.

This Is Just the Beginning!

The data highlighted in this article are just the beginning. A great deal of data were collected in the extended pilot. Additional data visualizations are available on our survey Data & Results page. Since we are still in an extended pilot phase, not all data have been released publicly, and school names have been anonymized. Identifiable school data will be available in future data releases.

In the next iteration of the Survey of Community College Outcomes, planned to be released in 2024, we will report data for each institution in the Fifth District. Please contact us with any thoughts or suggestions. We want to hear from you on how we can continue to make future improvements.