Until just a few years ago, colleges could anticipate classes of high-school graduates each bigger than the last. Two decades of steady supply drove enrollment growth and let campuses be choosy, gathering freshmen with good test scores and parents who could pay. But those days are over.
Peer into kindergarten classrooms across the country, and you will see fewer students. For every 100 18-year-olds nationally, there are only 95 4-year-olds. The Northeast and Midwest show the sharpest drop-offs, according to a Chronicle analysis. In less than a third of states, mainly in the West, can you find as many younger children as older ones.
Over all, fewer young children are white or black. In half of the states, more young children are Asian, and almost everywhere more are Hispanic. In the Southeast, at successively younger ages, the number of Hispanic children increases markedly.
College officials not already following such trends would do well to pay attention. Demographic change has become a crucial focus of enrollment management over the past decade, but with intense pressure to fill each year’s class, longer-term planning can lapse. And despite more data analysts in the field, much number crunching, experts say, is still clumsy. Campus leaders are often unaware of the coming reality, even as they rely on students for revenue.
Studying the pipeline is essential, says William T. Conley, vice president for enrollment management at Bucknell University. “If they weren’t born, they’re not going to go to college.”
Using data from the U.S. Census Bureau, The Chronicle examined by state and county the population from age 18, or zero years from traditional college age, down to age 4, or 14 years away. Younger age groups are strikingly smaller in New England, as in Rockingham County, N.H., where 18-year-olds number almost 4,500 and 4-year-olds just 2,600, a difference of more than 40 percent. With fewer young white children in almost every state, many counties’ younger age groups would be vastly smaller if not for much larger numbers of Hispanic children.
Take Gwinnett County, in suburban Atlanta, where more 18-year-olds are white (38 percent) than black (28 percent) or Hispanic (20 percent). But at age 7 and below, Hispanic children outnumber both other groups. Similar trends are apparent around the country. In Kane County, Ill., home to Aurora University, Hispanic children outnumber white children age 9 and younger. And in Marion County, Ore., home to Willamette University, the shares of white and Hispanic children in the youngest age group almost match.
The Chronicle’s analysis offers a snapshot of one point in time, not projections of the future. Numbers are of children, not students, and carrying forward the population of 4-year-olds, for instance, would not account for school enrollment, dropout, or migration. The measurement of years away from college age aligns with a traditional path, not those of the adults increasingly represented in postsecondary enrollment.
But as broad signals, the numbers show important shifts. Where younger age groups get bigger, people are poorer: Of about 450 counties with significantly more younger children than older ones, about 330 have median incomes below $50,000, compared with a median of $52,762 nationally.
By contrast, in many of the highest-income, most-educated counties—which have reliably delivered high-school graduates to colleges—the supply of younger children is dwindling. That pattern is striking in the suburbs of New York City: Long Island; Westchester County, N.Y.; Fairfield County, Conn. In Somerset and Morris Counties, N.J., both with median incomes of more than $98,000, the populations of 4-year-olds compared with 18-year-olds are 26 percent and 32 percent smaller, respectively.
Like a fisherman familiar with the best spots, a college recruiter might look to nearby Hunterdon County, N.J. It’s one of the wealthiest in the country, with a median income of $104,000. Nearly half of adults have bachelor’s degrees, compared with 28 percent nationally. The future recruiter’s problem? For every 10 older teens, Hunterdon has about five 4-year olds. In ponds that have long produced good catches, the fish are disappearing.
Colleges commonly cast about for students like the ones they already have: often affluent and white. Do these five ZIP codes north of Chicago generate the applicants you’re after? Try these three outside St. Louis. Define a go-to “geodemographic cluster” and look for similar profiles in other places.
These days, though, the chance of discovering a secret, stocked fishing hole is slim. The rate of population growth—0.7 in 2013—has not been this low since the 1930s. The overall fertility rate has gone down since 2007; for white, college-educated women, it stands below the replacement rate.
“There are virtually no major regions of the country where non-Hispanic white birth numbers are not getting lower,” says Steve H. Murdock, a professor of sociology at Rice University and a former director of the Census Bureau. “Basically, all groups besides Hispanics have birthrates lower than replacement.” If the number of white children seems to be increasing in certain areas, he says, it’s probably because families are moving there. (Several counties in The Chronicle‘s analysis with greater numbers of younger white children than older ones are home to major cities: Boston, Denver, Nashville, San Francisco.)
Those patterns put some colleges in a precarious position. Imagine a small private institution with an enrollment that is almost entirely white. “You are headed for a downturn,” says Mr. Murdock. “You’re likely to need to diversify in order to maintain enrollment.”
Jon Boeckenstedt knows the importance of data on that front. Now associate vice president for enrollment management at DePaul University, he has worked at five colleges in the past 30 years. Until recently, he says, tracking demographics was a frustrating task. In 1995 he tried to download a data set from the U.S. Department of Education’s Integrated Postsecondary Education Data System, or Ipeds, using a dial-up Internet connection and a 14.4-kilobit-per-second modem. “It ran about 23 hours,” he says, “and it died.”
Since then, access both to data and to software to display the data has changed the game, says Mr. Boeckenstedt. He can now download that Ipeds data set in 10 seconds, he says, and manipulate it in 20 minutes.
Many enrollment officials look to the Western Interstate Commission for Higher Education, known as Wiche, which publishes projections of high-school graduates about every four years. Some also examine—often with consultants—the Census Bureau’s American Community Survey and marketing databases like Nielsen Prizm, which identifies geodemographic segments such as exurban “Country Squires” and city-dwelling “Money & Brains.”
The last significant dip in the number of high-school graduates, in the 1980s, prompted many colleges to recruit more adult students, Mr. Boeckenstedt recalls. Now that market may be saturated, he says, and colleges will have to get more creative. A rare data whiz in his field, he analyzes and visualizes data on his blog Higher Ed Data Stories, in part to help colleagues navigate this wave of change.
“It’s not that uncommon that I’ll come home, and after dinner, when my wife is at yoga,” he says, “I’ll go on data.gov.”
Sophistication in student prospecting varies from one college to the next, but more of them—including public institutions, financially pressured into enrollment management—are seeking new methods.
In purchasing names of prospective students from the College Board and ACT, some enrollment officials are sorting the pool like baseball scouts applying sabermetrics. Analyses of current enrollment to guide recruiting can’t focus just on the bread-and-butter student, says Steve Kappler, assistant vice president and head of postsecondary strategy at ACT.
Given the growing population of high-school students who would be the first in their families to go to college, he says, one institution recently asked: Which first-generation students have done well here? It saw success among those with a high “interest-major fit,” according to ACT, meaning that the interests they reported on the test’s Student Profile Section matched their intended field of study.
The college now plans to recruit more first-generation students with a high interest-major fit and a high “mobility index,” a predictor of how far they’ll travel to enroll, which ACT added to its menu of characteristics in 2009.
Of course, if institutions recruit by buying names from the College Board and ACT, then only those who have taken a test, who are in some way preparing for college, are on the list.
In October, when Wiche published projections of high-school graduates for the 25 most populous metropolitan areas, it pointed out that some of the fastest-growing portions of the population, especially the males among them, are least likely to enroll in college. Brian T. Prescott, director of policy research at the organization, hopes the data will help cities and community groups better serve historically underrepresented students.
The White House is busy encouraging more outreach by colleges, which it says should recruit more low-income students by working with local school districts, for example, to mentor middle- and high-school students and bring them in for campus visits. Dozens of institutions made such commitments last week.
While gaps in college-going rates by race and ethnicity have narrowed in recent years, by income they’re still stark. In the high-school Class of 2012, 82 percent of graduates from high-income families were enrolled in college the following fall, along with 66 percent from middle-income families and 52 percent from low-income families. That 30-percentage-point difference is about the same as in 2002.
Now, with fewer children nationally, recruiting similar numbers may depend on raising college-going rates. Demographic shifts could change who gets treated like a prospective college student.
