As we approach the first full round of PRP awards in schools, and the shit well and truly starts to hit the fan, I feel it may be worth digging out an older piece I wrote in pre-blog days and dusting it off. Don’t say you haven’t been warned….
Performance related pay for teachers : “Inconceivable”
After fairly sustained pressure from Michael Gove, the STRB announced that a greater degree of performance related pay should be introduced for teachers. Who could disagree ? Surely teachers deserve to be richly rewarded when they are successful ? In trying to show you what might be coming, I’d like you to imagine the scene in the Princess Bride, in which the Man in Black (let’s call him the Dread Pirate Gove), challenges Vizzini, the Sicilian dwarf who kidnapped the princess. The challenge is that the Man in Black has poisoned one of two glasses of wine. Vizzini must choose a glass, and then both will drink together. The loser dies. The winner gets Princess Buttercup.
I’ll start slowly, as Vizzini did, while he mused over his options.
“There are two ways of measuring teacher “performance” : subjective and objective. The subjective method would see Heads and deputies grading teachers, usually on OFSTED criteria, based on lesson observations. There are two main drawbacks here : firstly, the quality of subjective judgements, and secondly, amount of time available for judgements. Despite what OFSTED would have us believe, there’s no clear relationship between OFSTED criteria, as subjectively applied, and results. It is entirely possible for a teacher to receive an outstanding grade from an observation, but achieve woeful exam results, and vice versa. And even if there were, then we hit the second issue, which is that there is no way that an SLT can realistically assess how a teacher is performing in this way. Usually, a teacher receives 1-3 hours of formal observations for performance purposes each year. Assuming most secondary teachers teach 25 hours a week, 38 weeks a year, then a maximum of 0.3% of their lessons would be observed. Not a lot of time in which to judge the efficacy of that teacher’s general output. Even a very weak teacher might be able to tick all the OFSTED boxes in three of his 930 lessons, while a consistently strong teacher (or one of her students) could have an unfortunately-timed bad day. I would happily bet my Department’s entire budget for next year that, were a subjective system of this nature be applied, then in the very first year, many schools up and down the country would see bonuses going to teachers with below-average results, while those with top results saw their pay frozen. It’s a racing certainty. So we clearly cannot choose the glass marked ‘subjective performance measurement’.”
“Aha”, says the Man in Black, “you choose the objective measurement. Then drink !”.
“Not so fast”, I say.
“Make performance dependent upon exam results, and you’ll have a clear jump in the pay of teachers of those subjects which are internally marked. Some subjects are 60% internally marked, or even 100% internally marked with some external moderation. You won’t be surprised to hear that those subjects often tend to obtain relatively strong results each year compared to more traditional subjects with 75-100% external marking. Moreover, studies have clearly demonstrated that students find it easier to get high grades in some subjects than in others. So measuring just by results would be clearly unfair to teachers of certain subjects.”
“Finished ?” asks the Man in Black, raising a disinterested eyebrow.
“Hang on, I’m just getting started ! Some subjects receive far more lesson time than others. For example, in many schools, maths, English and science receive significantly more curriculum time than other subjects because of their league table status. In my school, between Y7 and Y11, a student will receive an average of 2.2 lessons per week in history. Maths and English receive 4 lessons a week. So the little darlings have twice the teaching time before their maths GCSE than their history GCSE. Will I get a bonus if my results are only half as good as maths ?”
“You’re stalling now” says the Man in Black.
“You’d like to think so,” say I, “but there’s more !”
“Not all schools or departments can maintain the same teacher for the same classes. If a GCSE class obtains great results, do we reward the Y11 teacher who saw them to the exam or the Y10 teacher who saw them through the majority of the course ? What about the teachers who don’t have an exam class that year, or those who teach one paper each to the same A-level class? Do we break down marks by paper before awarding bonuses?”
The Man in Black sighs and looks bored. “You need to choose”, he says. “We can’t sit here all day.”
“No !” I cry. “You haven’t heard the half of it yet. Now let’s get serious : timetabling suddenly becomes determinant of pay. The English teacher gets her lessons periods 1 or 2 every day, with alert kids coming quietly out of formal assembly. The geography teacher gets her lessons as a double, on the last two periods of Friday, with tired students desperate for the weekend, following an art lesson with that scatty bloke who can’t control a class. Should we reward the English teacher for getting better results than the Geography teacher, or should we split the bonus between the English teacher and the member of SLT who timetabled the lessons ? Clearly, we cannot choose the glass marked ‘objective performance measurement’.”
“You’ve made your choice”, says the Man in Black.
“Of course I have,” I say. “As we’ve seen, a subjective system would certainly result in the perverse outcome of giving a performance-related pay award to teachers with worse results than others who go unrewarded, based on an unrepresentative sample of their teaching output according to criteria with no proven relevance to outcomes, so I can clearly not drink from the glass in front of you. But an objective system, as is completely obvious, could never be truly objective, and would be impossible to apply equally to all teachers as long as there are different subjects, different exam boards and even different times of the day. So I can clearly not drink from the glass in front of me.”
“You’re bluffing” says the Man in Black. “You don’t know which glass to choose.”
“Of course I do,” say I, “and it’s … oooh, look over there…”
In the real world, when I get hold of a class of students, then I like to think that I can have an impact on their results and their progress. I have a friendly rivalry with other Heads of Department which involves much ribbing on results days. I take great pride in trying to add as much value as I can, individually, and as head of my subject. However, I also recognise that when those students arrive in my classroom, their progress is also partially dependent upon others. For example, their English and RS teachers, whose approach to matters such as developing literacy and analytical thinking have a clear impact on their ability to negotiate the history course. Their results are also dependent upon the history teachers who taught them in previous years, giving them the basic tools for progress and instilling a love of history. They’re dependent upon when the lessons are timetabled and which lesson comes before they enter the classroom (and with which teacher). They’re also dependent on how I group the classes. A couple of difficult students can depress results for a whole class. Do I, as head of department, put those students in my class and hammer my own pay, or in one of my team’s classes to damage their pay instead ? The students’ results may also be dependent on whether I’m using the excellent resource produced by my colleague. After all, she now has a vested interest NOT to share that resource, in an attempt to ensure she gets better results than me so she can get the pay rise. Likewise, do I pop into school tomorrow and remove all those model essays I’ve written over the years, so that the other teachers can’t have their students benefit from them ?
No matter how schools try to implement this, they simply cannot, with any degree of confidence, isolate the impact of individual teachers from the myriad of other influences on a student’s achievements. If a school awards by results, then Governors should brace themselves for an endless procession of teacher appeals pointing out that they had fewer lessons/more difficult students/miserable timetabling/harder exams than the favoured few. If a school awards pay by SLT subjective gradings, then those Governors should brace themselves for a procession of teachers arriving with results clutched in their hands, asking why their stellar outcomes have gone unrewarded, despite being better than a newly enriched colleague’s results with the same students. Moreover, governors will have to ask themselves : do they want a school in which staff have a financial incentive to withdraw all co-operation from each other ? Where teacher is set against teacher, and resentment is fostered as unjustifiable pay differentials grow ?
In the Princess Bride, Vizzini sneakily switches the glasses while the Man in Black is distracted. They drink slowly. Then, Vizzini dies. The Man in Black admits to the princess that he had poisoned both glasses. But he was immune to the poison, so it didn’t matter to him that he’d introduced it into the wine. Whichever choice Vizzini made, was the wrong one.
Good luck in the pay round, everyone.
Life imitates art, and today I attended a meeting trying to set out how the PRP system will work this year in my school. Let me be clear, I don’t blame the various school managers trying to put something in place. They’ve been given their orders by Gove and Ofsted, and feel they have no choice. But there was a fantastic example of just how insanely stupid this whole system is.
We were told that all teachers had to set three targets, and one of those had to be an outcome target, one a CPD target and one a “leadership” target (there’s a “leadership” blog coming, by the way. It’ll be ugly). But for each target, teachers had to demonstrate the real impact of what they do. So if their target, the example went on, is to host a CPD session for their colleagues, then they have to demonstrate the “real impact” of that CPD on student outcomes.
Just think about that for a moment. Take a moment to consider that – on the face of it – reasonable statement, and then think about educational research. All the educational researchers, in all the universities in the world, have been trying to identify impact measures of different inputs for decades. And ultimately, they can’t even give such an impact measure for something as important as class size, or number of hours the student spends in the classroom, or the qualification level of the teacher. They occasionally say that they’ve identified some factor which they think has a positive or negative correlation, although they can’t quantify that difference. Then another university comes along and tells them that actually they’re wrong, as it’s really this variable which makes the difference.
So all these esteemed research professionals have proven unable to isolate and quantify the impact of individual inputs on student outcomes. Yet, under this amazing PRP system, 500,000 individual teachers in England are expected to quantify the exact impact of their one half-hour of powerpointed CPD on the outcomes of the students their colleagues teach. Or they don’t get a pay rise.
Generally, I try not to insult people (except Gove; and Wilshaw), but anyone who seriously supports performance related pay for teachers is a fool. That’s you, Policy Exchange idiots, and you, Gove, and you, Wilshaw, and maybe even you Tristram-Bloody-Pointless-Tory-Light-Hunt.
Hope that clears things up.