Digital divides considered harmless

Authors: Jonathan Grudin
Posted: Tue, April 21, 2015 - 8:00:16

The problem with early technology is that you get stuck with all this legacy sh*t.
– Director of Technology at a leading private high school

The impermanence of elevation differentials in seismically active terrain

Educational technologists have expressed concern about disadvantaged students falling farther behind; haves versus have-nots. Education faces challenges, but I assert that digital divides are not the big ones. Paradoxical as it may sound, divides can at times offset the advantages enjoyed by wealthier schools. The uninterrupted increase in the capability and decrease in the price of technology can be a challenge for early adopters. This is especially evident in education now, because primary and secondary public schools have reached a tipping point, but let’s first consider other domains in which apparent advantages proved to be short-lived or illusory.

Before Germany was reunified in 1990, wealthy West Germany had a strong technology infrastructure. It then invested two trillion euros in the former East Germany. Not all of the expenditures were wisely planned, but a strong digital infrastructure was. Soon after, a friend in the West complained that the East had better computational capability than he and his colleagues, who were saddled with older infrastructure and systems.

The exponential rise in capability and decline in cost often rewards late adopters. Who among us, contemplating a discretionary hardware upgrade, hasn’t wondered how much more we could get by waiting a few months? Early adopters spend money for the privilege of debugging a new technology and working out best practices through costly trial-and-error, after which the price drops. The pioneers establish roles and develop work habits that are shaped for systems that are soon surpassed in capability by offerings that may benefit from different approaches. The “have-nots” of yesterday who start today can adopt practices tuned to better, less expensive systems. They benefit from knowing what has and hasn’t worked.

In her 1979 book In the Name of Efficiency, Joan Greenbaum revealed that executives marketing early mainframe computers could not document productivity gains from the use of extremely expensive systems. Businesses paid millions of dollars for a computer that had roughly the computational power of your smartphone. Mainframe vendors were selling prestige: Through the mystique around technology, customers who bought computers impressed their customers, an indirect benefit as long as no one realized that the emperor wore no clothes.

This phenomenon was reflected in Nobel prize-winning economist Robert Solow’s comment in the mid-1980s: “You can see the computer age everywhere but in the productivity statistics.” It was labeled “the productivity paradox.” Two decades of computer use had delivered no apparent economic benefit.

Those who followed did benefit: purchasers of systems that arrived in the late 1980s and 1990s. The systems were much less expensive, and software designers had learned from the ordeals of mainframe users. Productivity gains were measured.

Optimistic technologists define the digital high ground. Their colleagues in marketing build dazzling if not always tangible castles, attracting those who can afford the price tag. Consider home automation. Fifteen years ago, wealthy homeowners built houses around broadband or tore up floors and walls to install it. This set them apart, but how great or long-lasting was their advantage? I soon heard groans about maintenance costs. Only after wireless provided the rest of us with equivalent capability at a small fraction of the cost did services materialize to make access valuable. The early explorers had to decide how long to maintain their legacy systems.

This introduces a second challenge: An aging explorer may not realize when the rapid movement of underlying tectonic plates shifts yesterday’s high ground to tomorrow’s low ground. Late entrants can steal a march on early adopters who are set in their ways.

To have and have not

Digital divides melt away. “Have-nots” become “haves,” and by then more stuff is worth having. In the 1970s, Xerox PARC built personal computers with tremendous capability, 10 years ahead of everyone else. The allure of working with them attracted researchers from minicomputer-oriented labs and elsewhere. It was said that for many years, no researcher left PARC voluntarily. In the 1980s, another research opportunity for the wealthy arose: LISP machines built by Symbolics, LMI, and Xerox, expensive computers with hardware optimized for the programming language favored for artificial intelligence.

There was a clear divide in the research community. And then, in the early 1990s, Moore’s law leveled it. High-volume chip producers Intel and Motorola outpaced low-volume hardware shops. I remember the shock when it was announced that Common LISP ran faster on a Mac than on a LISP machine. LISP machines were doomed. Less predictably, interest in LISP (and AI) declined, perhaps due to a loss of the mystique that masked a failure to deliver measurable benefit. PARC researchers had made landmark contributions, albeit not many that Xerox profited from, but as its researchers shifted to commercial platforms, PARC’s edge faded. A digital divide evaporated.

The same phenomenon unfolded more broadly. Through the 90s, leading industry and university research labs could afford more disk storage, networking, and high-end machines. It made a difference. A decade later, good networking was widely accessible, storage costs plummeted, and someone working at home or in a dorm room with a new moderately priced high-end machine had as good an environment as many an elite lab researcher with a three-year-old machine. Money still enables researchers to explore exotic hardware domains, but for many pursuits, someone with modest resources is not disadvantaged.

The big enchilada

Discussions of haves and have-nots often focus on emerging countries: the challenges of getting power, IT support, and networking. I thought harvested solar energy would solve the problem sooner than it has. Mobile phones arrived first. In many emerging regions, phone access is surprisingly close to universal. Soon all phones will be smart. If mobile, cloud-based computing is the future, those in emerging markets who focus now on exploiting mobile technology could outpace us, just as Germans in the East leapfrogged many in the West.

Promoting accessibility to useful technology is undeniably a good thing. But our optimism about our wonderful inventions can exaggerate our estimates of the harm done to those who don’t rush in. Some of the same people who lament digital divides turn around to decry harmful effects of the over-absorption in technology use around them.

Education: And the first one now will later be last…

For over forty years I didn’t think computers were a great investment for primary or secondary schools, even though my destiny changed in high school when I taught myself to program on a computer at a nearby college that was unused on weekends. It sat in a glass-walled air-conditioned room and had far less computational power than a hand-calculator did twenty years later. I first programmed it to discover twin primes and deal random bridge hands. It was fun, but I saw no vocational path or educational value—the college students weren’t using it, my classmates had other concerns, and maintaining one cost more than a teacher’s salary. Pedagogy was the top priority in K-12. I may have sensed even then that ongoing professional development for teachers was second. As the decades passed and costs declined, having a computer or two around for students to explore seemed fine, but a digital divide in education didn’t seem a threat. Some wealthy schools struggled with expensive, low-capability technology, subsidizing the collective effort to figure out how to make good use of it.

But the times they are a-changing. It was evident two years ago. New pedagogical approaches, often called 21st-century skills and tied to Common Core State Standards: critical thinking, problem-solving, communication and collaboration skills, adaptive learning, project-based learning, and adaptive online-only assessment. I’ve seen this fresh, intelligent reorientation in my daughters’ classes. Students and teachers face a learning curve. Parents can’t help much, not having experienced anything like it. But if successful, the results will be impressive.

Software will be useful in supporting this. Sales of high-functionality devices to schools are rising rapidly. The price of a good tablet has halved in two years and will continue to drop. Public K-12 schools are joining the private schools that are 1:1 (“one to one,” each student carries a device everywhere). When not poorly implemented, 1:1 is transformational. I have attended public school classes with beneficial 1:1 Kindle, iPad, Chromebook, and laptop PC deployments.

Students in the past used school computers when and how a teacher dictates. A student issued a device decides when and how it is used, in negotiation with teachers and parents. The difference in engagement can be remarkable. Third and fourth graders are strikingly adept, middle school seems a sweet spot, and high school students are often out in front of their teachers.

Hardware and software value propositions change dramatically with a 1:1 deployment. When a student can take notes in all classes, at home, and on field trips, good note-taking software is invaluable. A good digital pen shifts from being an easily lost curiosity, when used once a day in a computer lab, to a tool used throughout the day to take notes, write equations, sketch diagrams and timelines, adorn essays with artwork, label maps, and unleash creativity.

Dramatic changes in pedagogy are enabled. Time is saved and collaboration opportunities are created when all work is digital. Teachers who formerly saw a student’s work product now see the work process.

Rapidly dropping prices and recognition of real benefits are bringing 1:1 to public schools in many countries, erasing a digital divide that separated them from elite private schools that could previously afford to issue every student a device. Of course, it requires preparation: Professional development for teachers, addressing new pedagogical approaches and technology, and wireless infrastructure for schools, which is expensive although also declining in cost. Most teachers today have experience with technology. The shift to student responsibility and initiative helps—tech-savvy fourth graders need little assistance when they reach middle school. In fact, students often help teachers get going.

As 1:1 spreads, pioneers risk being left behind. Private-school students are often from computer-using homes where parents have strong preferences for Macs, PCs, or Android. As a result, many private schools adopted a bring-your-own-device (BYOD) policy. Teachers who face classes with myriad operating systems, display sizes, and browser choices are limited in assigning apps, giving advice, and trouble-shooting. They are driven to lowest-common-denominator web-based approaches. In the past, students nevertheless gained experience with technology. A digital divide existed, though the benefits of technology use were not always evident.

Public schools can’t require parents to buy devices. Fewer public school students arrive with strong preferences and almost all are delighted to be given a device. School districts get discounts for quantity purchases of one device, and teachers can do much more with the resulting uniform environment. When preparing to go 1:1, one of the largest U.S. school districts had classes try over a dozen different devices in structured tasks. In summarizing what they learned, the Director of Innovative Learning began, “One of the things that our teachers said, over and over again, is, ‘Don’t give students a different device than you give us.’ That was an Aha! moment for us” [1].

Many public schools that carefully research their options choose this path. The future may be device diversity, but teachers struggle with non-uniformity. They can’t assume students have a good digital pen, which is far more useful when available for sketching, taking notes, writing equations, annotating maps, and so on in every class, at home, and on field trips. Bring Your Own is rarely the way to start learning anything. Instruction in automobile mechanics began with everyone looking at the same car to learn the basic parts and their functions. Allegory is taught by having a class read Animal Farm together, and later encouraging independent reading. Digital technology is no different.

I have visited many public and private 1:1 schools. Some private schools that are BYOD do not realize how much the affordances have shifted. Back when technology capabilities were limited, their students were on the advantaged side of a digital divide. Today, technology can make a bigger difference, and I have seen public school students who benefit more than nearby private school students.


Serious economic inequalities affect healthcare, housing, and nutrition. Before adding technology to the list, consider its unique nature: a steadily flowing fountain of highly promoted but untested novelty that takes time to mature as prices drop. We nod at the concept of the “technology hype cycle.” We are not surprised by productivity paradoxes. Yet some books that belabor technology for its shortcomings and frivolous distractions also decry digital divides: Questionable technologies are not available to everyone! The news is not so bad. Courtesy of Moore’s law and those who are improving technology, divides are being erased faster than they are being created.


1. Ryan Imbriale. The smart way to roll out 1-to-1 in a large district. Webinar presented March 11, 2015.

Thanks to Steve Sawyer, John King, and John Newsom for observations, comments, and suggestions.

Posted in: on Tue, April 21, 2015 - 8:00:16

Jonathan Grudin

Jonathan Grudin works on support for education at Microsoft. Access these and related papers at under Prototype Systems.
View All Jonathan Grudin's Posts

Post Comment

No Comments Found