There is a (possibly apocryphal) story from the days of the Raj, when, worried about the number of cobras in Delhi, the British government announced a bounty program to reduce their numbers.

At first it worked and numbers went down. But one thing humans are good at is making the most of circumstances, and soon local entrepreneurs went into the business of breeding cobras so they had a steady supply of serpentine revenue. This meant that they no longer had to catch any wild cobras and numbers went back up.

The Brits, getting wise to this wily con, mutinied on the bounty. At which point the breeders released all their cobras back into the wild and their numbers shot up higher than before: this became known as the Cobra Effect.

This is a form of the law of unintended consequences: where a solution to a problem could actually make it worse.

Similar to this is the Hydra Effect, after the eponymous Greek cryptofauna - you cut off one head but two more spring up. Schools are full of cobras and hydras. A plan, best laid at Inset, gangs aft agley.

The most common version of this is when a new initiative is launched in a school with little consideration of whether the intended participants actually have time to execute it.

Data monitoring, paperwork and bureaucracy are the chief culprits, when some bright spark announces that from now on, to be able to evidence progress, all teachers will have to triple mark, or fill out feedback forms every week, or keep records on dragon skin or something.

Harried staff try to meet these witless deadlines, and sacrifice time better used in planning or teaching. Everything suffers, and nothing is done properly. Tempers rise, morale falls, along with teaching quality. Soon, even the dragons stop turning up.

More cobras slither through behaviour management.

Take exclusions. No one likes to see pupils excluded either internally or externally. And many schools would - correctly - love to see their exclusion rates go down.

In fact who can blame them, when your average inspector will frown on any school with significant exclusion stats? Any school dreading the lash and cosh of Ofsted will fret about publicly available data that draws attention to itself.

One pseudo-solution some schools pursue is just to stop excluding pupils, which certainly solves one problem, but creates others more serious.

Exclusions serve many purposes, from punitive to restorative, but one function they have is to temporarily reduce pressure on a class community trying to deal with isolated, extreme disruption. When pupils aren’t removed even when the school process indicates that they should, then classrooms can struggle.

There are many very good reasons why a pupil presenting challenging behaviour may need to be removed from a class - for support, for example. With an artificially deflated rate on internal exclusion, the figures look good to an inspector, but the actual classroom culture suffers, and so does the pupil who needs support.

Another problem this causes is that every behaviour system is a continuum, and that process must have a terminal point.

Consider a pupil who refuses to work because they don’t like the teacher’s ethnicity. The pupil may start to incur sanctions - detentions, loss of privileges etc. But what if the pupil persists in their non-compliance? Or their misbehaviour escalates?

Suppose the pupil couldn’t give a stuff for sincere and scaffolded support. Suppose they refuse to attend detentions or conversations?

You would expect an escalation of the school process, and the complexity of the response. But what if all that doesn’t work? What then?

There must be a point at which the school must be able to say: “This pupil’s behaviour exceeds our capacity to amend, and they need to taught in alternative provision of some form.”

And if this option is unavailable because the school wants to keep the cobra figures down, a tremendous disservice is done to the whole system.

The school process backs up, as pupils realise that, with persistence, they can defy school conventions with impunity….and nothing will really happen to them.

That knowledge is corrosive. It is the universal solvent for a school’s culture. It implies that everything is permitted because nothing is truly forbidden.

We may not like excluding pupils, internally, externally, permanent or fixed term, but they are a necessary part of the system.

The desire to reduce exclusions by simply turning off the tap ironically creates circumstances where their use is required more and more, as misbehaviour backs up the pipe and remodels the social norms of the school in the direction of incivility and belligerence.

Another variant of this is the brilliantly named Streisand Effect, after the famous chanteuse. In 2003, she objected to a Kenneth Adelman, a photographer who had, as part of a series of photographs charting erosion around the Malibu coastline, included a picture of her fabulous (of course) clifftop mansion.

Her lawyers chased Adelman and his website with a $50 million lawsuit. Word spread about this legal homage to sledgehammers and walnuts, and where before the image had been downloaded six times, the site subsequently received in excess of 420,000 visits. 

Streisand found - like Ryan Giggs and Jeremy Clarkson and many more since - that “don’t look at this!” is a terrible strategy to stop people looking at something.

I once sat in a staff meeting after summer break. The school’s results had tanked and everyone knew it. We all sat there waiting for guidance about what to do next.

But the leadership didn’t mention it once, and instead we were treated to inspirational videos and Inset/death by sugar paper. An elephant was shaking it like a Polaroid picture through the room, but we weren’t allowed to mention it.

It was the managerial version of shouting “do not think of a lemon” at people. All we could think about were lemons.

Education is full of unintended consequences, mainly because we often focus more on what we’d like than what we’ll get.

We’d like happier students, so we play them a video. But then learning becomes movie night and we wonder why the grades went backwards.

Or we want the school grades to improve, so we enter them for exams we know they’ll pass. And then we wonder why our curriculum resembles a Happy Meal.

Or you want to dazzle Ofsted, so you bury staff in marking policies that Darth Vader would think was a bit demanding. And then wonder why everyone wants to leave.

But thinking any intervention a) only has one effect and b) that effect is predictable and linear, is as sensible as assuming that you can stop getting older by banning birthday parties.

And sometimes, if you want fewer cobras, you don’t start by just targeting cobras. You need to create conditions where cobras don’t multiply.

Tom Bennett was a teacher in the East End of London for 10 years. He is now the government’s behaviour tsar, and the director and founder of ResearchED, a grassroots, teacher-led project that aims to make teachers research-literate

Want to keep up with the latest education news and opinion? Follow TES on Twitter and like TES on Facebook