The company was pitching a promising but unproven technology to some of the worst non-point-source polluters in California. These operators couldn't control stockyard waste runoff, chemical drainage from olive packing, and many other kinds of dispersed pollutants that didn't come out of a pipe and that thus were impossible to filter. Our company was pitching something new to a conservative tightfisted industry--agriculture--and it had limited money and not very good connections in that industry. So how could they find customers while they were improving the actual technology?
The answer had several parts.
(1) Carefully select technical results for the sales pitch to potential customers. The chemists looked for results from streams that were "most like" the one they were being asked to treat, but this also meant exaggerating positive results with the best of intentions, based on the commercial chemist or engineer's "we'll fix that tomorrow" optimism that really does solve a lot of hard problems sooner rather than later.
(2) Claim a finished product for a work in progress. The phrase was "this machine can be dropped off the truck" and was "plug and play." That this was not true was an open secret. The customers knew that the machines had to be monitored and tweaked with enormous human input during every day of operation. But "plug and play" was a validating fiction that reassured all parties that everything was moving in the right direction.
(3) Use (1) and (2) to obtain a steady series of trial installations of the new technology at problem sites. To do this, the company bid below cost to get in the door, both to get upfront market share and to accumulate technical experience that could improve the product. The company lost money doing this, but gained visibility, reputation, and know-how that brought its exaggerated claims onto a glide-path toward reality.
(1) plus (2) plus (3) equaled ongoing financial losses. This was a start-up company with a high burn rate of mostly angel funding. In the two years during which I was associated with this particular firm, they never achieved accurate estimates of "normal" long-term operating costs. This wasn't because they were dishonest, but because there were no "normal" operations in this entry / development / sales / start-up mode.
The most important part of the process was (4), moving regulation towards the existing technology.
The company hired Washington and Sacramento lawyers with personal ties to legislators, including a couple of pro-green Senators, and to officials in organizations like the EPA. At the time, non-point-source pollution was a growing regulatory concern--a major hurricane had just spread hog farm waste over much of eastern North Carolina--so many politicians and senior agency managers were interested in this company's work. On one trip, the company also met with representatives of large engineering firms. Their project VP laid out the market penetration process: we get you further installations, we collect results together, we validate your results, we explain the results to officials, and we help insert the results into the details of a policy's enactment. Getting into a market meant creating a market of which you the company were already the center.
Until I attended several meetings like this, I had assumed a difference process, which might be called "how a finding becomes a rule." I thought regulators tracked findings of scientists that were studying an important pollution problem, and then wrote pollution standards that used formal study findings to define correct health and environmental goals. Science and regulatory standards would come first, and then the public and private sectors would partner to meet regulatory standards.
I learned in 1999 that the reality was the reverse. Technology of commercial interest to a firm was presented to regulators via brokerage through lawyer-lobbyists. The technology would help define the scientific parameters of the pollution problem. Since all of the firm's capital had been raised by promising angels and other investors a major share of a "trillion dollar market" in global water treatment, the company presented only those technological solutions which offered the company potential market share, meaning technology over which they had a monopoly via a combination of patents and trade secrets, mostly the latter.
The private firm's business plan was blended with appraisals from third parties and regulatory intentions from public agencies. The agencies, ideally, would recognize the company's results as valid and then write the standards to fit the company's technology. The explicit and sanctioned goal is sometimes called "regulatory capture" -- regulations would conform to the strengths of a potentially important technology, which was also the technology controlled by the company in question.
Today, MOOC leaders clearly understand the process of adapting state regulation and social needs to the technology they happen to control. They are proceeding systematically through the steps above:
(1) Sales momentum has to stay ahead of the accumulation of negative product results. Sometimes results are misstated. Coursera's page about its "pedagogical foundations" links to the famous Department of Education meta-analysis of about 45 studies of online education's effectiveness. Coursera says, "This analysis demonstrates very convincingly that online learning methods are, on average, at least as effective as face-to-face learning."
In reality, the study comes to different conclusions. It finds that "the studies in this meta-analysis do not demonstrate that online learning is superior as a medium" (52). It suggests that the benefits of "blended" online / face to face techniques may flow not from the technology but from the fact that they "involve more learning time, additional instructional resources, and course elements that encourage interactions among learners." The most cited online study to date raises the intriguing possibility that the benefits of the only good kind of online--relatively expensive "blended" modes--don't actually come from "online" (p. xiv), but from instructors and students having more time and money.
Sometimes results are simply ignored. For example, a new, high-quality study of 500,000 online courses in the state of Washington (Xu and Jaggars 2013) found,
Overall, the online format had a significantly negative relationship with both course persistence and course grade, indicating that the typical student had difficulty adapting to online courses. While this negative sign remained consistent across all subgroups, the size of the negative coefficient varied significantly across subgroups. Specifically, we found that males, Black students, and students with lower levels of academic preparation experienced significantly stronger negative coefficients for online learning compared with their counterparts, in terms of both course persistence and course grade. . . . These patterns also suggest that performance gaps between key demographic groups already observed in face-to-face classrooms (e.g., gaps between male and female students, and gaps between White and ethnic minority students) are exacerbated in online courses. This is troubling from an equity perspective. (23)Online may not only not improve educational outcomes. It may worsen them. It may take a higher ed system that has for years been worsening race and class inequalities and make them more terrible still. But in a start-up sales regime, momentum depends on minimizing or ignoring findings that contradict the company line.
(2) The distinctive feature of MOOC marketing in 2013 is the shift from being an exciting experiment to being offered as a working solution to budgetary and access crises. Hence the legislation proposed last week by the California State Senate Pro Tem Darrell Steinberg simply presumes that online courses could put "quality first" and offer expanded access with no real loss of educational value. MOOC leaders will themselves note various weaknesses and unrested assumptions in their approach, as several did at UC Berkeley's Learning Mode conference this weekend (#learningmode). The flaws in MOOC quality are open secrets, but they are subsumed by plug-and-play claims embedded in the rhetoric of Big Data and related tropes of inevitable technological progress. This powerful combination swamps both the history and the current reality of on-line limitations, some of which harken back to the correspondence schools of the late 19th century, which 130 years ago also promised increased individual attention and the bypassing of the "sage on the stage."
The Steinberg legislation marks the synthesis of MOOC steps (3) and (4), in which large scale trials are being insured through a state-created artifical product market revolving around Udacity and Coursera in particular. The business problem is this: Large-scale trials must be had at any cost, or the product momentum will die, investors will have doubts, money will dry up, market penetration will fail. MOOCs have shown that lots of people will sign up for a free online course--and that a tiny proportion actually persist. If students are required to pay tuition, as with UC online, they currently don't sign up in the first place.
Thus 2013 may not be Year of the MOOC II, in that it may reveal that MOOCs may have no large natural market of tuition-paying students. To head off this possibility, the firms have shifted focus to regulatory capture. This is what happened when Udacity was hired by San Jose State University to run 3 remedial courses. The formal signing ceremony put founder Sebastian Thrun on the same level as the governor of the state and the chancellor of the Cal State University system.
The bigger payout of this regulatory lobbying strategy appeared last week, when California's public college and university community learned in The New York Times that state legislation "would force colleges to honor online classes" ("force" has been altered to "seeks" in the current online title, h/t Wendy). The article appeared a few hours after I happened to hear a report about Udacity founder Sebastian Thrun leaving the capitol offices of State Senate Pro Tem President Darrell Steinberg, the author of the legislation. There are other signs of the kind of regulatory capture I witnessed in my corner of 1990s start-up culture. Ry Rivard reports on the secrecy that surrounded the writing of SB 520's amendments, which do force all three higher education segments to accept online courses approved for credit by a separate state-established panel. The fact sheet calls for "a thoughtful and strategic harnessing of Silicon Valley's innovations in online education"--signalling the skewed history typical of lobby capture.
MOOC momentum is being driven not by educational need or proven technological achievement but by a business lobby with connections and resources as good as Wall Street's, and with a better social cause. The movement's systematic exaggerations, the lack of concern for impacts on public university ecosystem, the staged benevolence towards a hostile customer--all are hallmarks not of technical or pegagodical progress but of a carefully designed business strategy. We won't be able to assess the technology correctly unless we see this.