Wagging the Long Tail
Editor's Note: This article was corrected 21 Dec, 2022, to correct the earnings number and add the source citation.
The term “Long Tail” gets bandied about a lot among authors. Understanding how it applies to authors, earnings, revenue, and books can help you make sense of some of the statistical nonsense that surfaces periodically.
Statistically speaking, the long tail is a property of distributions. When the highest data values stack up close to one side and fall off rapidly as they're plotted out, the distribution of the data is considered a long tail.
Normal distributions – the famous bell curve – flatten on each end with the majority of data values stacked around the middle. We're most familiar with these and use a lot of measures of central tendency when describing them. The mean or average usually represents the peak of the hump. The standard deviation tells you something how clumped up the data is. Data points too far away from the mean become outliers and sometimes get ignored except in footnotes.
For a long tailed distribution, those measures of central tendency aren't meaningful because there's no center. The value at the point on the curve where the number of data points above is the same as the number below – the median – provides a better perspective. The two colored regions in the plot (one green, the other yellow) above show where the median for that dataset occurs. Each region has exactly the same number of data points.
Plotting annual author income generates a long tailed distribution. The same can be said of units sold and revenue earned. They're all long tailed distributions because they're all derived from the same basic measurements about the sales of individual books, just aggregated in different clumps.
A few books sell the most units, earning the most in revenue. The authors who wrote those books will earn the most income. Everybody else will share the remaining sales, revenue, and derived earnings. The individual authors may swap places based on exactly what data you use, but they'll all still be there somewhere.
Learning that £7,000 (just over US$8000 at time of writing) represents the median annual author income[1] tells us something important. For every author who earned less than £7,000, there's another one who earned more. Sure, a lot of authors earned next to nothing, but for every author who earned less than the median, there's one who earned more.
The big question that nobody seems to want to deal with logically is how to raise that median value.
You can't just pay authors more.
Well, you can but it matters which authors you pay. It doesn't change the median point if exactly the same number of authors earn more than the median as before, no matter how much more those above the median make.
It doesn't matter if you offer higher advances. They get paid this year, but what about next when they can't get the next book out? They're just swapping places in the line, not making any progress. It doesn't help if those larger advances don't earn out and the author has to start over. They can't earn out if the bookstores don't stock the books. Narrow bookstore margins don't allow bookstores to take many chances. They need to bet on the surest winners.
Result: Publishers can't afford to publish books from people they don't know, in niches that are too small, or genres bookstores consider passé. Bookstores won't order those books and too many of the orders get returned.
To raise the median, you'd have to pay more to the authors who earned less than the median, the authors that don't sell very well. They'd drag the median up with them, but where does that money come from?
Self-published books follow the same long-tail. The tail is a lot longer. More people make some money. The median is going to be a lot lower than $8000. A lot of self-published authors earn next to nothing. I dare say more people make at least “side-hustle” level incomes there than with trad, but it's only because there are so many more of them. For self-pub, the long tail doesn't get truncated by the finite limits of traditional production and distribution.
The crux of the issue remains: Authors get paid when their stories sell.
You can't pay an author more unless they sell more stories. Publishing more books helps, but you can't sell more stories if readers aren't buying them.
Regardless of publication route, the distribution guarantees only a few authors will earn the most money, a lot of authors will earn some, and a lot more authors will earn nothing – or next to it. No matter how you plot the curve, the median value only shows that exactly half the authors earned more and half earned less.
Now if we could get readers to buy more books, there might be a solution.
For more discussion of the ACLS/CREATe survey, see: A Rising Tide
Next Up: Disrupt This
- CREATe. (2022) “UK Author Earnings and Contract Report.” https://www.create.ac.uk/wp-content/uploads/2022/12/Authors-earnings-report-DEF.pdf