Canada is still organized around the agricultural calendar, but the crop -- other than peaches and tomatoes -- that signifies fall to me is the harvest of new books that starts appearing around Labour Day. By now authors are already traipsing across the country to hawk their wares at literary festivals, bookstore shelves are laden with promising titles, review sections are burgeoning and press conferences are being organized to announce the lucky nominees for the Giller and the Governor-General's awards.

All of this activity, although it is as predictable as the leaves turning colour, is the result of choices made on the basis of whims, educated guesses, gambles, instinct and taste. Who reads at festivals, who gets reviewed, who gets nominated -- all those decisions are made by individuals, acting alone or on committees. In a word, the process is highly subjective.

What is supposed to be clinical, however, are the cold hard data determining which books make it on to bestseller lists. Those lists are supposed to reflect what books consumers are actually buying -- not reading, necessarily, but certainly what they are purchasing and bothering to take home with them. If you lived in Britain, Australia or the United States, you could be fairly certain that the title listed at the top of the bestseller list actually is selling better than the next one down on the list. That's because booksellers in these countries use a standardized computer tracking system called BookScan that collects and reports point-of-sale data. There is no such national tracking and reporting system in Canada linking publishers, wholesalers, distributors and booksellers.

In this country, bestseller lists are compiled in such an idiosyncratic and subjective manner that Globe Books editor Martin Levin referred to them as the BS lists in Saturday's paper. "Years ago," he wrote of the Globe's bestseller list, "we devised a system that . . . intends to reflect what our readers might actually want to buy. Like every other BS list in existence, it is partial and only functionally accurate."

For its part, Quill and Quire, the publishing industry newsletter, bases its bestseller list on a survey of approximately 160 independent bookstores in various parts of the country. The rankings do not reflect sales in Indigo or in stores such as Wal-Mart or Costco -- in other words the majority of booksales in this country.

"We are in the Dark Ages," says publisher Kim McArthur, who has been sitting on an industry-wide committee studying the way suppliers and retailers in other industries and in other countries track sales and exchange electronic data. "Have you noticed how when a movie opens, we know how many people went the first weekend?", she asks rhetorically. "What we do in books is to say, 'Let's hold our finger up in the air and guess how many people bought our books over the weekend.' That would never, ever happen in a grocery store, in the movies or in the record industry."

While other countries have geared up electronically, Canada has been "mulling it over," according to McArthur and she, for one, is very frustrated to have another fall season go by without a system in place. Part of the problem, and part of the solution, is that the Canadian approach involves a public/private partnership with the federal government.

Instead of just signing on to BookScan, the system pioneered and owned by ratings conglomerate A.C. Nielsen, the Canadian way is to set up committees, study the alternatives, provide funding to enable the various players in the complicated supply chain that links writers and readers to upgrade their computer facilities and finally set up an arms-length agency to administer the process. In the end, the agency may well decide to go with the A.C. Nielsen system, but that decision will not be made until there has been a great deal of sober reflection and deliberation.

The delay is frustrating, but there is a lot more at stake than simply being able to say with authority which book really is No. 1 on the fiction bestseller lists. Once in place, a comprehensive tracking system would eliminate potentially costly mistakes such as the time, back in the fall of 2000, when Chapters ordered an additional 10,000 copies of Margaret Atwood's The Blind Assassin from McClelland & Stewart. Before rushing back to press, M&S did some hurried sleuthing and found 15,000 copies of the novel sitting unnoticed in a corner of the chain's huge warehouse. If M&S had printed and shipped the books, they would have probably been returned to them at the end of the season for credit.

That is one of the reasons the government is involved in the steering committee. "Taxpayers invest about $50-million a year into this industry," says a senior official in the Department of Canadian Heritage. "Depending on whom you ask, returns are running at 30 to 40 per cent, and 80 per cent of all returns are pulped," he said. With figures like that, it doesn't take a task force, even in a country like Canada, to conclude that taxpayers' money is not being used efficiently. "We would rather be spending money creating new books," agreed the official "and promoting new writers than on books that are going to be pulped."

Still, governments by nature are in the consultation business and that process rivals molasses when it comes to speedy action. No system will work effectively unless everybody is compiling bibliographic information in the same way and using compatible electronic communications software, argues the federal bureaucrat. Otherwise, it will be "garbage in and garbage out." In the end, we will probably have a much better and more comprehensive system that will benefit the industry, and ultimately taxpayers.

There is, however, a potential downside. At the moment, we like to boast that our bestseller lists are much more literary than, say, those in the United States, which are almost always topped by bodice-busters and thrillers. Once we give up the fantasies about the books we think people should be reading and substitute the books they are actually buying, our bestseller lists will probably look a lot more like the American lists. They will be more accurate, but we might not like what they say about our vaunted literary culture.
smartin@globeandmail.ca