Last year, Macmillan was fined over eleven million pounds in the UK for corruption in East and West African educational publishing. Essentially, they bribed officials to get their books preferred over competitors’ books. A year later, Oxford University Press has been fined for similar practices, also in East Africa.

In South Africa today, we have a major textbook crisis in Limpopo and worrying signs of textbook meltdown in the Eastern Cape. Last year’s astonishing decision by government to approve only eight textbooks in any one subject area may have increased the temptation to pay off officials. Big publishers are reeling after government rejected several major titles for public school use. Small publishers are even bigger losers, where rejections effectively put them out of business. Legal action by one publisher in response may have serious consequences for the rollout of the new curriculum for grades 11 and 12 next year.

How broken must a business model be before we stop perpetuating it?

The problem starts with the way we’ve structured schoolbook-publishing business models. This is an industry driven by well-meaning people. No one puts up with the poor salaries and crazy pressure of schoolbook publishing if they don’t care first and foremost about good education. So the system is broken not because of malice or incompetence. It’s broken because, fundamentally, the way we sell textbooks is flawed. And today, technology is showing up those flaws more clearly than ever before, and provides alternatives.

The two fundamental flaws are these:

1. Government has to approve textbooks before teachers in state schools can use them.
2. Textbooks are priced and sold individually. As a downstream result, middlemen handle distribution, adding costs and bungling deliveries.

We will never be free of corruption and failed distribution as long as these two flaws exist.

If you have ten minutes to read on, I’ll explain why.

Why government approval is bad

The first objection to government textbook pre-approval is usually that centralised decision-making is inefficient and undemocratic. Others will argue that centralised approval is crucial, because it prevents poor books from messing up education, and that government simply has to get better at vetting books. Both arguments are strong, but neither addresses the whole problem.

If we are serious about holding teachers and schools publicly accountable for their performance (an issue hotly debated in the US), we must let them choose their own tools, and expect them to get better at choosing over time. They must then have the widest possible number of choices. The state can’t limit the materials teachers and learners can choose to learn with.

Textbooks will then get more diverse and, at the high end of the spectrum, better and more creative. Diversity in any system lets better products evolve, and poorer ones disappear. And by ‘better’ I don’t mean more colourful, or more glossy, or more technologically advanced. I mean better in that they meet the needs of their customers better, needs that only end-users themselves can know. Right now, government requirements result in a one-size-fits-all textbook model. When I was publishing schoolbooks a few years ago, I found it impossibly difficult to differentiate my books from our competitors’ under the narrow constraints of curriculum documents and what we expected from state approval panels. And yet my potential target markets were utterly diverse: city and rural, English, Xhosa, Zulu, Sotho and more, wealthy and poor, progressive and conservative, religious and agnostic, visual and textual, hearing and deaf, sighted and blind, and young and old for their grade.

In a recent round of state approvals, some leading publishers’ books were rejected because somewhere in the book they gave away the identity of the publisher – something that’s supposed to be hidden to prevent corrupt panelists favouring or disfavouring a given publisher’s book. What a pity. A publisher’s identity should be so clear from the nature of the book – and panelists so knowledgeable about textbooks – that hiding it is impossible. If publishing companies are supposed to be full of creative educators (something they all strive to be), they should each be bringing big, bold, brand-specific features to their books. I should know Macmillan’s books from MML’s or OUP’s at a hundred paces.

Instead, government approval panels leave us with the same, safe material from every publisher. Publishers can’t push boundaries, experiment, or offer completely new ways of presenting material. Without clear differentiators, publishers must compete only on the relative efficiency with which they can produce more than one title in a subject (sometimes under completely different brands to give the appearance of choice), manage the clumsy and expensive textbook-submissions and marketing process, and make smart political choices – for instance, in whom they choose as authors. None of these areas makes for better books. All are to some degree corruptible.

A few months ago, the Department of Basic Education printed four million open-licensed textbooks developed by Siyavula (see ‘A sea-change in South African schoolbook publishing‘). The deal was criticised by publishers for being a kind of state publishing, in part because the Siyavula books had been rejected, in an earlier draft form, by the DBE’s own approval panels. The irony is that, by not pandering to standard government requirements while developing their books, Siyavula had been more independent of state control than conventional publishers had ever been. The fact that the DBE decided to print Siyavula books for millions of learners only shows that the DBE themselves are not wedded to approval panels when other approaches to content distribution are available.

Approval panels won’t go away tomorrow. But the more we try alternatives, like the Siyavula deal, the more we’ll see organisations adapting textbook-making business models to suit them. For instance, corporate-social-responsibility budgets could fund the development of branded, open-licensed textbooks produced by small teams of pros – and that government might distribute them like they did the Siyavula books.

The question is: will existing publishing companies be among the first or the last to try those business models? Or will they stick to what they know: pricing and selling books one copy at a time?

Why pricing and selling per copy is bad

The second flaw in the industry is that publishers set prices and sell their books per copy. They do this because it seems intuitive and has always been done that way. To get a fixed unit cost per copy, you add up your investment in content development and predict (by any measure you can) how many copies you’ll sell. Slap a margin on that for a rough guide to profitability. Then you hold thumbs hoping you’ll beat the system and sell more than your sales estimate, boosting profits with economies of scale at the printer.

Unfortunately, apart from guessing wildly at a future-sales estimate, you’ve also planted the seeds of corruption and inefficiency. Single-copy pricing is only possible when the entire production and distribution chain are closely controlled. Economically speaking, you have to manage the scarcity of a physical product to keep prices where you want them.

To do this over a large area, you and government appoint middlemen – mainly booksellers and distributors who pitch or tender for a piece of the action. It’s not the existence of middlemen that are the problem. The problem is that they’re specially appointed. When only certain people get to profit from being middlemen in a chain, they will be sure to take a pound of flesh from it, sometimes by whatever means necessary.

Publishers and government have to get out of the scarcity-controlling, middleman-appointing game. If they didn’t have to control an entire distribution chain, they wouldn’t have to appoint special middlemen, and the potential for corruption and inefficiency would fall dramatically.

So how would this work? The state still needs publishers to make books, and will either pay for that to happen or encourage an ecosystem where publishers can get paid by others. Payment should be determined not by the price of each copy, but by what the parties agree is the value in currency of a given project. Instead of debating copy prices that are based on fictional sales projections, they would be negotiating a single, large licence or funding deal. Textbook development might be funded by the state (in lieu of buying finished copies of books as they do today) or paid for by businesses who fund textbooks as a branding exercise. Imagine “Vodafone Grade 8 Technology”, funded by Vodafone and open-licensed.

Once a deal is done, the publisher would release the content. Ideally, it would be entirely open-licensed, so that anyone could distribute it. The state could print it, individuals could put it on their phones and mail it to each other, copy shops could make and sell copies. Publishers could even allow remixing, so that others could build their own businesses or reputations on remixing the work (giving credit to the source publisher), making the publisher’s future source material that much more valuable to the education ecosystem, and therefore to funders.

Most importantly, there would be no pre-appointed middlemen controlling distribution and marking it up. There would be, instead, an open market for any business – including the publisher – to deliver content to learners and teachers, and to help them use it.

(A second prize could be a licence deal allowing distribution to a specific region or institution. That also requires managing scarcity – publishers or the state would have to take action where others outside that region or institution use the material in violation of the licence. But having been paid for the initial deal already, the publisher’s losses are less urgent. And while they decide how to act, the world’s better off for the material getting to those who need it.)

Each teacher or learner could spend a budget, or even their own money, on learning materials they choose (where this budget sits is immaterial, as long as individual teachers and their learners get to make the final decision about what to spend it on, incentivised by a range of independently devised, public performance scores). Those teachers and learners, spread in their thousands over towns and cities and villages, would get better and better at choosing what works best for them. Whether that’s booklets from a small time NGO with an old photocopier (perhaps using a service like Paperight), glossy textbooks from an old-school publisher, or mobile web resources from a multinational tech company.

And as a publisher, once everyone’s passing your books around, you can start to build other business models around that. For instance, you can sell support services around assessment and training and computer support. This is exactly what Siyavula is doing, and what Pearson already does in some markets. There is no reason to think that it won’t be as profitable as selling by the copy.

Working like this, our industry could be more transparent, and we’d see proper competition among distributors large and small. Most importantly, learners might actually get better textbooks, anywhere, on time.