New NCES Data Reveals Insight Into Online Market Demand by State and Region

By Senior Analyst, Brian Fleming

The recent release of NCES’s Enrollment in Distance Education Courses, by State: Fall 2012 has led to more conversation about the use of federally reported enrollment metrics to make sense of the size and impact of online learning. Of particular interest to us has been the introduction of state-by-state enrollment metrics, which are immensely valuable for estimating regional concentrations of online learning activity for determining pockets of demand and market saturation and for mapping activity to regions in which we find correlation with key economic indicators or market dynamics. This type of information is incredibly useful for late-adopters looking to leapfrog into an already mainstream market and unaware of the challenges of national reach, as well as long-time national providers seeking a more strategic regional focus.

What This Data Tells Us

With a better sense of how online learning has taken shape regionally, we can now begin to explore:

• What regional markets are more active, and why.
• If demand for online programming correlates with higher or lower economic growth, educational attainment, or income-levels in a particular region.
• Why particular types of online programs, whether discipline-specific or by modality, are more popular in some areas and less in others.
• The value proposition of for-profit education and why these types of institutions may succeed in certain markets and not in others.
• The balance of opportunity for seizing market share in areas of high demand while navigating the risks of entry into areas of high saturation.

NCES Offers Supply, Not Demand-Sided Metrics

To be clear, NCES does not report where students taking distance education courses themselves are located but where the institutions enrolling these students are located. This “supply-sided” rather than “demand-sided” framework can be misleading. For instance, at face value and without much thought, we could easily conclude that Arizona, for instance, home to two of the largest distance education provides in the country has an unusually high concentration of students learning at a distance (360,835, in fact), which is untrue. Arizona has a high concentration of institutions enrolling students online, but interestingly enough a lower concentration of students in this State enrolled in online programs. Where students choose to enroll and where they actually live are two entirely different questions altogether as far as NCES is concerned.

NCES has also only recently begun collecting distance education metrics and promises to include far more robust data sets in the future, likely to include student location, but does this largely for policy and regulatory purposes only, aimed most recently, we suspect, to help Federal and state regulators map cross-border activity by institution to determine what institutions are enrolling the most students out-of-state.

The good news is, given the need to analyze cross-border activity, NCES also chose to include a small but significant set of metrics related to the percent of students enrolled in distance courses that reside in the same state as the institution, out-of-state, and even internationally. This forms a useful backdrop for a demand-sided analysis of regional concentration of online enrollments, as below.

How We Do It

Based on regional and in-state enrollment intensity, and following our most recent headcount estimates of students learning online (defined as programming in which 80% or more of study conducted at a distance and limited only to degree-granting institutions), by institution, we calculate a regional concentration of demand-sided estimates based on a careful triangulation of relevant data points.

• Sampling from NCES state-level data. Because institutions report the relative % of students located in- and out-of-state, we start with analysis of 1,746 institutions reporting 50% or more of their total distance education enrollments coming from in-state. This leads to a total sample of 1.4 million students, nearly 45% of the total market, and a likely picture of enrollment activity. Looking then at the location of these institutions, we see, for instance, that many of those reporting 50% or more in-state also fall into narrower categories as high as 90% or more in-state. Many are also located near state borders, some of which, we determine based on a scan of websites and marketing materials, offer in-state/flat tuition rates geared toward recruiting students living adjacent to or near state borders. Other points that come into view include data on states with highly restrictive policies regulating out-of-state provides (e.g. Georgia and Alabama) and states that prioritize online education through public university systems and consortia (e.g. California and Maryland), both of which fuel higher in-state concentration.

• Data on consumer demand. We then test our estimates with our national survey of adult learners, which offers a glimpse of preferences by modality and data based on the location of each respondent. Based on a sample 5,325 adults interested in going back to school, we see that the highest preference for “Wholly Online” or “Mostly Online” programming comes from adults located in the Southeast, Midwest, and Southwest respectively. The Northeast has the lowest preference overall for “Wholly Online” programming. We also see that, based on a sample of 7,642 adults responding to our survey, 75% would prefer to enroll in an online program with a campus nearby, which when correlated with the location of the respondent helps further inform our understanding of where online learning is likely to be most concentrated.

• A market taxonomy. Finally, we take into account the intense consolidation of the market as a whole, namely that about 3% of institutions offering online programs enroll 45% of the total national headcount.This enables us to quickly make sense of what large-scale providers have the broadest national reach, namely Phoenix and Ashford, and which are more regional, namely University of Maryland-University College and a large number of community colleges such as Ivy Tech. Removing those with the highest national reach, and assuming these types of institutions already seek out enrollments in areas of high demand, from our overall geographic estimates we find that 75% of the total market headcount among those with less than 5,000 learning online originate from the same state as the institution anyway. With the location of these institutions in view, we then manually scan where available, public data from each of these institutions, analyze, and test assumptions of the relative demand for online by region to arrive at a reasonable estimate of where online students themselves reside, by region.