Does Online Reinforce the Color Line?

Ten years into the current round of austerity in public higher education, state funding levels remain 17 percent below their per-student levels of 25 years ago and 15 percent below their fiscal year 2008 level, adjusted for inflation. Public colleges and universities have on average managed to make up for these cuts with tuition increases: in adjusted dollars, net tuition has doubled since 1991 and is up by one-third since 2008. And yet in the process we have reduced public confidence, social impacts and student expectations.

Nonetheless, politicians often tell university officials that previous levels of funding will never return. Just as often, they explain that universities don’t really need the missing funds because of the money-saving role and educational benefits of new educational technology. This solution has re-emerged in this budget year. Our example is California, where we are again hearing claims that online will soon be as good as face-to-face instruction, and therefore public colleges can do more teaching with less money.

In January, Governor Jerry Brown proposed giving the California State University and the University of California even less than the inadequate 4 percent the systems thought they were getting -- a nominal 3 percent, 2.1 percent after subtracting one-time money. Nor did the governor propose a tuition increase to compensate for a state allocation that barely meets consumer price inflation. The two university systems have 33 campuses between them, and Brown’s budget insures the continuation of structural problems and everyday squeezes at all 33.

Why do the governor and most of the Democratic establishment think UC and CSU can do more with less, and that state cuts don't hurt quality? In a press conference following the release of his 2018-19 budget, Brown offered a kind of explanation.

It is enough. You're getting 3 percent more and that's it. They're not going to get any more. And they've got to manage. I think they need a little more scrutiny over how they are spending things. It's just because the university is a good they say “we've got to have more good” -- but if you have too much good it -- in certain circumstances -- it becomes a bad. So they're going to have to live within their means. And what will happen here is when the next recession [comes] they'll have to put everything in reverse and lay people off and raise tuition and that's not a good thing. So, they've got to lower the cost structure and there are tools to do that and they need to step up and more creatively engage in the process of making education more affordable.

We leave aside the standard charge of administrative bloat. Bloat is real but arises in large part from efforts to replace funds lost to public cuts: staffing grows to manage expanded fund-raising, industry and foundation partnerships, customer-friendly student services, continuous publicity and brand management, and real estate development and other auxiliary enterprises. Bloat functions here as a mode of finger-pointing rather than reform.

More importantly, Brown categorically assumes that money can be saved by shifting face-to-face instruction to online. Last year, he and California Community Colleges (CCC) chancellor Eloy Ortiz Oakley, whom Brown had also appointed to the UC Board of Regents, announced a fully online CCC degree program called the Flex Learning Option for Workers (FLOW). His budget proposes that this become a new “online campus” for the community colleges. The idea is similar to what then dean of Berkeley Law Chris Edley suggested for the University of California almost 10 years ago -- an 11th UC campus that would be all online. The background assumption has remained the same: ed tech has moved the cost-quality curve, so online college is both better and cheaper than face-to-face.

True believers in online cost miracles were tested toward the end of 2013, when the massive open online course wave crashed on revelations that MOOCs were neither better nor cheaper. The famous Brown-brokered deal between Udacity and San Jose State was suspended after an National Science Foundation-based study showed Udacity's online courses actually reduced remedial education outcomes, prompting Udacity founder Sebastian Thrun to call MOOCs a lousy product.

In addition, those of us who tried to find cost savings were unable to. Georgia Tech's online computer science master’s program -- Udacity's flagship program -- continues to run with multimillion-dollar subsidies from AT&T. (Also see the ambiguities of the University of Florida's online programs.) In short, online had not suspended the economics of education, in which institutions lower costs by lowering quality. After 2013, MOOC companies retreated from their initial claim to be replacing college, instead offering professional retraining and credentialing.

In focusing on adult re-trainers, Brown's current online proposal seems at first to follow the arc of retrenchment. But it comes with a renewed MOOC-style claim that online is "as good or better" than face-to-face. George Skelton quotes CCC chancellor Oakley, making a categorical assertion of the online program's value because its technology is directed at "social network kids."

A further example appears in Teresa Watanabe's coverage in the Los Angeles Times.

In Watanabe’s article, Laura Hope, a California Community Colleges executive vice chancellor, said improved classes and tools for online orientation, counseling and tutoring have significantly narrowed the performance gap between online and traditional classes. Nearly two-thirds of online students completed their courses in 2015-16, compared with just over half a decade earlier. Over the same period, the percentage of students who completed traditional classes stayed roughly the same, at about 71 percent.

There is no doubt that distance learning can be more convenient than traditional, face-to-face course taking. This convenience likely explains the growth of online learning over the years. Looking at data available from the California Community College Chancellor’s Office (CCCCO), in the 2006-07 academic year 68,546 full-time-equivalent students (FTES) enrolled in for-credit distance education (DE) course work. By the 2016-17 academic year, this number grew to 164,966, an increase of roughly 140 percent. As a proportion of total for-credit community college full-time-equivalent enrollment, distance education has grown from 6.6 percent in 2006-07 to 14.7 percent in 2016-17. In contrast, for-credit nondistance education shrank by nearly 2.5 percent over this same period.

Convenience granted, the real question is educational: Is online "as good or better" than face-to-face, as good or better in a way that justifies using online courses to replace face-to-face instruction or to launch an entirely online community college campus? CCC's own data analysis is now making this claim, summarized in this slide, which refers to success rates in all types of CCC courses for the past 10 years.

In the email that accompanied this figure, a CCC official claimed it showed that distance education success rates are on track to match the results of face-to-face. So, via CCC, "better" is back. And "cheaper" (though dubious and unproven) never went away.

When this figure appeared, we were working on a paper about what the underlying CCC data said about racial disparity in online courses. Our overall question has been, does moving students of color from face-to-face to online help or hurt their education? In other words, how does online learning impact success rates across racial groups?

We were able to reproduce the CCC-provided figure above with the publicly available data we've been using (give or take a percentage point). However, we had been disaggregating outcomes by type of course and by racial category. Here are two examples of face-to-face/online comparisons, using two types of courses that are likely to resemble the "retraining" courses offered by the FLOW program.

As Figure 2 illustrates, distance education continues to deliver a significant drop in success rates in basic skills courses. The convergence trend CCC claimed on aggregate is much weaker here. In addition, online makes the racial disparity of in-person courses somewhat worse. The success rates of underrepresented minority students, to use the standard classification, are poor. In addition, they are not improving much in basic skills, as CCC claims for the aggregated results.

One reasonable policy conclusion would be quite the opposite of Brown's and Oakley's. Black and Latino basic-skills students should never be placed in online courses -- not until researchers are given the time (and data) to explore and overcome the mechanisms underlying the racial disparities. And even though white and Asian students outperform the state averages, a face-to-face/online gap also exists for these student groups. Consequently, it seems reasonable that even they should use them sparingly.

Brown and Oakley would likely push back against the use of basic skills distance education course success rates as projections for future success rates of students enrolled in the FLOW program. This is a fair point if, in fact, FLOW will draw the kinds of nontraditional, older, working students the program intends to. But will the FLOW program really only serve students who are “college ready” and not in need of some developmental education?

So far it is unclear. But let’s assume that students in the FLOW program will resemble current “vocational” education students. These students are, in the aggregate, often older, more career focused and arguably more reflective of the FLOW program target demographic.

Figure 3 shows that success rates are generally higher in vocational courses relative to other course types (e.g., basic skills). This is true with online as well, where the convergence with in-person is more convincing. Yet racial disparities remain, and remain larger than with in-person courses. Again, a reasonable conclusion would be that a system that is serving a majority-minority student population needs to be very conservative with its use of these courses.

All of this leads us to advance the following points:

1. Online education is valuable and important as a selective and supplemental approach to extending in-person higher ed. It helps students who cannot stop full-time work or family care. It is especially good at dealing with the repetition that is part of all learning. This is an area where it has a clear advantage over human teachers, as language labs (and books!) have been proving for generations.
2. State leaders are wrong to continue to push online education as a categorical value. This current push depends on aggregating data in a way that conceals how online disadvantages African-American and Latino students. Online education is currently an engine of racial inequality. No good higher ed policy can be created by ignoring that fact.
3. It is still unclear how learning in online environments compares to learning in traditional environments. The chancellor's office seeks to justify the growth of distance education by pointing to improvements in the success rates of students in these courses. But by reducing “success” to completion, the CCCCO masks the differential impact of DE course taking on traditional measures of successful education, including both cognitive (e.g., learning, retention) and noncognitive (e.g., interpersonal skills and attitudes) growth. It is doubtful that the prospective employers in the state will hold the same definition of success as the CCCCO. Consequently, we recommend a rigorous, longitudinal evaluation of the state’s extant online courses prior to the new construction of an entirely online campus.
4. Online education should never be used to excuse state budgets that are too small to support the established features of educational quality. These features include the presence of fully qualified teachers working with classes that are small enough to allow individual feedback. Online courses that approach the quality of face-to-face courses are actually "hybrid" courses that involve structured personal contact. We know of no hybrid online courses that will save universities money. States should never budget by assuming the opposite.

In short, university officials, including faculty senates, should loudly oppose officials who let online reinforce the color line. The FLOW program should restart the discussion about the higher ed practices and investments that would actually reduce racial disparities in attainment, rather than cover them up.