I was intrigued last November by this Marginal Revolution post on why music reviews tend to be positive, compared to film reviews. I was particularly interested in a comment comparing film and book reviews:
Book reviews are generally positive because reviewers frequently have a choice of which book of several to review, and choose to read books they expect to like, and then to give publicity to ones they enjoyed (knowing that there’s no such thing as bad publicity). It seems like there is a three-step process:
1. Is this book likely to be worth my time?
2. Read it. Is this book worth writing about?
3. Write a review.
Movie reviews basically have to cover all releases in a week, so there is no such filtering out of bad products.
My perception is that book reviews tend to be softer than movie reviews, but I had a different theory: that for social reasons people in the relatively small literary community are reluctant to give negative reviews to people they are likely to meet, if they don’t know them already. As the vast majority of reviewed films are foreign, this is less of a problem in movie reviewing.
A friend who is a part-time literary critic made a related point, that many book reviewers are actual or aspiring book writers themselves, while few movie reviewers have made or are likely to make a film. There are several reasons why this may lead book authors to be softer reviewers: they don’t want to provoke negative reviews of their own work, compared to a critic less of their reputation with readers relies on providing good advice to book buyers (which could lead to lower quality of reviewing overall, not just in being soft), and having been through the pain of writing themselves may just feel sorry for authors, even if the book isn’t much good.
Having an empirical mindset, I decided to try to test these theories by monitoring film and book reviews in the Weekend Australian, the Sydney Morning Herald, and At the Movies (deleting David Stratton overlaps). I have done this since 17 January, concluding this weekend. This gives me a sample of 105 book reviews and 60 film reviews, enough for some initial testing of the theories.
An initial methodological problem was that while film reviews have a convenient 5 star system, this practice is not used for book reviews (why not? another curious difference between the two reviewing systems). So I had to attribute stars based on my reading of the review. When the superlatives were flowing I gave 4 or above, when the review was broadly positive but not enthusiastic I gave 3 or 3.5, when it was lukewarm I gave 2.5, and when the reviewer clearly did not like the book I gave it two or below. Mostly this wasn’t too hard to do, but occasionally reviewing practices such as barely mentioning the book under review while discussing the general topic of the book made this difficult. It is hard to know whether this is a tactic to avoid giving a negative review, or the reviewer is so vain that he believes we would rather hear his views on the topic than the author’s.
The results do show the pattern I expected, but not nearly as strongly. On average, book reviews averaged 3.5 and movie reviews averaged 3.4. Among the book reviews, people who were writers themselves averaged 3.6, while people I categorised as critics averaged 3.4 (I could not classify all reviewers into either category, and classified people as critics where I knew they were regular reviewers, their biographical note said they were reviewers, or they were academics).
At the Movies averaged a 3.3 review, while the newspapers which run fewer reviews averaged 3.5 – consistent with the theory that the more reviews you have to run each week the more likely you are to cover weaker films. The SMH, for example, reviews only two films a week in its Saturday edition, and gave no ‘fail’ marks and only one 2.5 review – absurdly in my view to Clint Eastwood’s excellent Gran Torino, which they probably thought would be good.
To do this properly would require a much larger sample and at least two people starring the book reviews. But overall it suggests to me that despite some structural issues that count against book reviewing being conducted in the interests of book readers, the profession is not as corrupted by structural problems as I thought it might be.