Following on the thread started by others (thanks John and Rowena), it's not only who would be a writer, but who could be (and who could keep being one) and why. You see, anyone can write. And anyone can keep writing. Being published however, is another matter. That's about numbers sold. Now there are examples of publishers ignoring numbers (I know one of an author getting into the amazon top 100, to have the publisher take her book out of print, and my publisher, Baen, showing faith in an author with lousy numbers, which was well-rewarded later.) but they're rare, as rare as stories of publishers accepting that not every crash is a driver error (or every win driver skill) There are occassional stories of a publisher admitting the cover sucked, and trying again. It happens and much credit to those with the courage to do this, but generally bad numbers... are bad retail orders. It's not even that your publisher hates you, or that no-one actually knows a bad cover, bad distribution, poor laydown can sink a good book just as a great cover and great promotion and spend (with the good distro and laydown this creates) can sometimes launch a very mediocre one to success. Of course even that can fail, and that really is driver error. But bad numbers = bad odor clinging to that author's computer record = bad retail orders next time = death of the author's career. Now, as Louise pointed out, there can be utter insanity applied to retail orders -- and those are still your numbers, nothing you can do about it. (Which affect at worst your career, and at best the re-order status of your book. Low numbers put your 'name' into the re-order on demand category, not reorder on sale category of chain retail), and sadly these numbers do principally get dictacted by the chains. To put it comparible fishy terms (for me!), it's like basing all your fishery quotas on a situation where the fishery is dominated by a couple of companies selling fishmeal - which they catch by purse seine of schools of anchovy on the surface. This would lead you to conclude demersal fish - such as hake - are very rare. In actuality the purse seined fish are overfished and the bottom dwellers abundant -- but that's not how it looks. The numbers -without mathematical correction for factors such as cover, targeting, marketing, distribution, sellthrough (ie. a measure of re-order and distrubtion-cover) - are the equivalent of estimating gut-length by measuring turds. It is true that short guts cannot produce long specimens, but the converse isn't true. None-the-less, this is the situation right now and new authors need good numbers from chain retail. Which leads me to offer advice I detest (because I blame chain retail for a lot of the problems in publishing and reading, and I am a staunch indie supporter): If you are having a book come out, generate as much public interest as possible, and where possible get your audience to pre-order from chains, as this will affect their initial order numbers, and very possibly the re-order status.
That was distasteful, but had to be said.