Imagine a world with very few women. They barely ever talk to one another, and when they do it’s only to talk about men. Doesn’t really bear thinking about, does it?

Unfortunately, you don’t have to imagine this- films of all stripes have already shown it to you.

A seemingly easy test – the Bechdel test – highlights this. It has only three rules: the film

1. has to have at least two [named] women in it
2. who talk to each other
3. about something besides a man

My friend Chris Mead recently wrote about this here, detailing how few films pass, and why this is so troubling.

It would be foolhardy to dismiss the significance, especially considering that if we were to reverse the genders (I.E.- a film must have at least two male characters who talk about something besides a woman) so many films would pass as to render the question pointless. It does, however, leave something to be desired in terms of value assessment. Consider the following two examples:

If two female characters talk in the most stereotypical way possible about shoes, the film will pass.

If a female character talks about particle physics, but happens to be talking to a man, the film will fail.

So it’s quite possible to mount an argument pro or con if motivated enough. I’m personally much more interested in answering two questions: why is the pass rate so low (and gender representation in film so poor), and can we form a more satisfying model?

Why so low?

Naturally, when considering why representation on screen is so lacking, my mind turns to representation in the industry. This year’s Creative Skillset Census indicated a step forward for women in film, representing 46% of the workforce (thanks in large part to advocacy groups including Women In Film & TV and Birds Eye View), and 46% of creative development. While this is encouraging, there’s still far to go, and the figure representing the entire creative sector is just 38%. The development figure also doesn’t reflect above-the-line decision making, as it includes lower-level positions such as Script Reader. Looking at the BFI independent film industry stats 2010-12, just 16.1% of all writers of independent films released during the period were female.

It seems obvious that films written by female screenwriters are much more likely to pass the test than those written by male ones, but can this be backed up with hard data?

If we look at films included on the Bechdel Test Movie List, and take each fiction feature released this year, 76 of 182 pass uncontested, giving us a base level of 41.7%.

Focusing just on those which credits include at least one female writer, 32 of 49 pass uncontested, raising the pass rate to 65.3%.

Focusing further still on those which credits include only female writers, 13 of 19 pass uncontested, raising the pass rate again to 68.4%.

(Including contested passes, these figures are 62.6%, 85.7% and 89.5% respectively.)

Granted, the more we focus on female writers, the smaller the sample becomes, and the harder it is to generalise from. This, however, only upholds my earlier point- gender representation in screenwriting is so wanting, we can’t even cite a decent sample.

Towards a better model

The trouble with the test to many is that it looks at representation in sheerly quantitative terms- it’s just about the number of female characters present (and talking to each-other), rather than their quality as characters. It would be easy to alter a screenplay to pass without adding any value, simply by arbitrarily adding female characters. This is essentially the Smurfette Principle at work.

The test’s detractors say we should look at the quality of female characters alone. This too is problematic. If Smurfette* were made a well-rounded character, that wouldn’t alter the fact that she’s the only female character, which doesn’t sound like any town I know. True representation needs both number and quality.

I’ll now suggest an alternative test. We’ll call it the Jules Test (for the sake of gender ambiguity, ahem), and see if we can address these concerns.

First, we’ll look at quality. List all the prominent, memorable characters in the film. This will be difficult and will depend on interpretation. You didn’t expect this to be easy, did you? To cheat, start from the front page of the film’s IMDb listing and edit accordingly. We’re trying to isolate the characters who time and effort have been put into; the ones that are both distinctive and believably human.

Now comes the number analysis. Split your list of characters by gender, and work out the percentages (that’s the number of the gender you’re working on divided by the total number, multiplied by one hundred).

If the percentage of distinctive female characters lies within 40-60% of the total, that’s a base-level pass. If they lie within 45-55% that’s a merit and a smiley-face sticker.

Why the upper limit? Well, the goal is to be representative of society. If your film contains a 100% female cast, I’d have to commend you for redressing the balance, but I’d be lying if I called it representative. True, it may represent a specific subculture or situation (much like the all male ’12 Angry Men’ represents life inside a jury room of its time), but these have to be considered statistical outliers.

Yes, this is woefully subjective and difficult to apply. However, as you may have guessed, I’m not really talking about existing films here. I’m talking about your film. Look at your screenplay. Does it pass the Bechdel Test? Does it pass mine? If you can’t answer ‘yes’ to either of these questions, it’s time to do some rewriting.

* I should point out in fairness that both recent Smurfs films pass the Bechdel test, the principle being named after the television series.