Fixing accountability means doing less accounting – and finding better ways to account for our actions.
What’s wrong with accountability through accounting?
Accountability through accounting means measuring, quantifying and reporting:
- What proportion of students are currently meeting their target grade?
- How many students are receiving interventions?
- What proportion of wallspace is covered by appropriate posters?
Mostly, it’s a farce.
As a fairly new teacher, I reported termly how much my students had learned, using National Curriculum sub-levels and sub-levels.
But my school wanted more accurate, up-to-the-minute data. So they broke sub-levels into sub-sub-levels (5a+, 5a, 5a-, etc.). And they moved from termly to half-termly reporting.
After each data input, the head of department wrote a report. Then they met with the head to agree what action was needed.
Memorably, this once led the head of department to check why I had the worst-performing Year 8 class. ‘Worst-performing’ turned out to mean two students below their expected levels. Gemma was below expected levels because she didn’t come to school – I’d never met her. Mia was below expected levels because she’d spent half the preceding term in hospital.
This multi-step charade produced information I could have offered initially – if anyone had asked.
I guess this seemed precise and powerful, on some plane. But the assessment levels we used didn’t mean anything. (Level 6 meant make links between causes. Level 5 meant begin to make links. Presumably Level 5c meant beginning to begin to make links. Level 5c+ meant whatever you wanted it to mean.) And no one checked that what I’d called a ‘5c+’ actually deserved that.
It didn’t take me long to realise I could save a lot of time and hassle by teaching as best I could – and reporting data… generously. (Protesting, by claiming Gemma had gone up a level when I’d still never met her, went down less well.) My bosses reached similar conclusions I guess, when they invited us to reconsider the levels we’d reported for Year 9 before they were reported to the DfE, as they currently came below national averages.
So much for accountability through accounting. Usually, it means we:
- Submit regular (and often time-consuming) reports
- Report quantitative data (reflectingly only narrowly what we know and do)
- Are checked for compliance (but not for accuracy)
- Use measures which can easily be gamed (and soon are)
Things change. I’m sure many schools have better systems. But I doubt they solve all these problems. You can use GCSE grades instead of levels, but they still tell an incomplete tale. You can moderate teachers’ grading, but only partially, and at a cost in time. You can mistrust teachers’ reporting, but you need someone who knows better – and is happy to burn a few relationships.
I guess this is why rigorous reviews of this kind of accountability in schools consistently show that it doesn’t increase student learning.
Getting rid of accountability… may not be the dream
If accountability isn’t working, perhaps we should get rid of it. Why can’t we trust teachers as professionals? Why can’t we be more like Sweden and Finland, who manage perfectly well without inspections and tests? There are only two barriers:
Barrier 1: Accountability matters
When we’re not accountable, things go wrong. Dan Ariely wrote a great book full of experiments showing few people will rob you blind, but most will round things up in their favour (2013). Nemesis stalks Hubris, Ariely himself has been caught publishing research based on fake data. You’ll have to decide whether that strengthens or weakens his case.
When people aren’t held accountable for their actions, things fall apart. Never you, dear reader. Other teachers though, graded Teacher Assessed Grades very generously. More vividly, consider the chaos which ensues whenever we lack parking wardens. Or recall that it only took a twelve-hour police strike for a prosperous city to see “Fires, explosions, assaults and a full-pitched gun-battle.” We may believe in humans’ innate goodness, but it depends on accountability for our actions.
Problem 2: Leaders need to know what’s happening
If leaders are to do anything useful, they have to know what’s going on. Imagine being a headteacher, or, God forbid, running multiple schools, and trying to make decisions without some knowing what’s happening in your five, or fifty, or five hundred classrooms. What to change? Who to help? How? You might wish leaders just left you alone (one in six teachers do). But they won’t, they can’t – and keep their job. And an uninformed leader won’t stop leading – they’ll just take uninformed decisions.
Which is why, when you scratch the surface, you find that fairy stories about Nordic countries are usually just that. No, Finland doesn’t have an Ofsted. But no, schools are evaluated, and there are national tests, and they matter (more here). Yes, teachers decide grades in Sweden. And yes, people take advantage of that – particularly privately-run school chains which “systematically set higher grades than public schools.”
We can’t avoid accountability: instead, we need to make it better.
Accountability through giving account
As a teacher, the crazy thing about Ofsted – and most observations – is that you rarely get to account for your actions. Sometimes, you know a student needs constant support. Sometimes you know they can succeed independently – despite asking for help. I defy a passing observer to judge your decision without asking you why you made it.
We can all object to entering meaningless data: accountability through accounting. We can’t reasonably object to describing, justifying and discussing our actions and choices. This is accountability through giving an account. Wherever you stand ideologically, I suspect you agree that you owe some account of your actions to at least one of students, parents, colleagues, leaders, taxpayers, the community, politicians and posterity. (For this distinction between account and accountability, see in particular Andrews et al., 2017, Ch.5.)
Accountability through account can be a powerful influence. Supreme court judges cannot be gainsaid. The split-second decision making of special forces cannot be second-guessed. But in both peer groups, peers’ expectations and judgements hold them to account (Ibid.).
To encourage accountability through giving an account, we might:
- Support groups of teachers to agree and articulate standards: these are examples of effective questions in our subject
- Help peers and leaders carefully examine existing practice: we’re going to look closely at questioning this week
- Look to listen and learn before passing judgement, considering both thin data (how many Grade 5s?) and thick context (how did Period 5 feel?): “Why did you choose that question for Alfie then?”
At a classroom level this feels like some kind of lesson study. We might look at student books. We might ask teachers to explain their plans, and the choices they’d made. We might review exit tickets together, and discuss what could be done differently next lesson. We’d want to balance genuinely challenging our approach and our thinking – with genuinely taking account of what’s really happening, and why. I suspect we’d learn more; I suspect it would feel fairer.)
(Yes, many leaders do this. My argument is that this – not reporting data – should be the core of our accountability processes.)
Pity the leader who gives up on accounting
For a leader used to getting clear data showing the performance of every student every half term, this sounds scary. How can we aggregate such nuanced, idiosyncratic, individual data.
I think it nudges leaders to think less about accounting, and more about the ingredients which would allow them to hold teachers to account:
- Models – how do those teaching (or running departments or schools) know what good looks like?
- Processes – how do those teaching (or etc.) teach/support/observe well?
- Recruitment – how do we get the best possible people teaching (or etc.)?
It’s much harder. The ‘data’ you get is much messier. But it better reflects what’s really happening. And by a twist of fate, it might actually help improve teaching and keep teachers happy – in a way accounting can never do.
The obvious concern is a descent into chumminess. Martin knows that Michael’s department results are poor, his attempts to help his colleagues weak, his prospects dismal. But Martin also knows that Michael gets his rounds him, and (connectedly), his wife just left him. So Martin accepts Michael’s contention that his department is still ‘Good’ – even though he knows inside it doesn’t quite make the grade.
There are two ways to check this. First, we can keep some kind of accounting – as part of this account (Honig and Pritchard, 2019). I’m not saying don’t use student tests, I’m saying don’t rely on them. A discrepancy between the teacher’s account and the evidence of student learning would encourage closer examination.
Second, we can use external assessors in the account-giving process. This can get chummy too. But I’m reminded of the Imperial Chinese legal system, which accorded huge power to the district magistrate. Officials were chosen from among those who had performed “extraordinarily well” on the Imperial exams. But censors – who inspected magistrates – were chosen from those “who did even better (Friedman et al., 2019, p.10).” The external assessor shouldn’t just make up the numbers: they should be the best teacher, head of department or headteacher we can find.
Conclusions
1. Accounting is omnipresent, but it’s rarely enjoyable, rarely accurate, and always dispiriting.
2. But getting rid of accountability creates more problems than it solves.
In its place:
3. We need to think more about how we can help teachers/leaders/schools account for their choices.
4. This means listening and seeking to understand – without whitewashing poor practice.
So, we can ask:
5. What useless, time-consuming and inaccurate accounting procedure can we cut? (If you’re not sure, ask your team. They’ll know.)
6. What models and processes can we put in their place?
If you liked this, you might enjoy…
I recently wrote about the quest for certainty in schools, and why we should get used to acting on incomplete data.
Less recently, I suggested better ways to use data in schools, inspired by Moneyball.
Even less recently, I wrote about measuring what we value, not valuing what we measure – and showed how it worked in the classroom.
I learned about the account/accounting distinction from the work of Matt Andrews and Lant Pritchett, which I can’t recommend highly enough. Start with Chapter 5 of this.
I’m working to get leaders better information about what their staff are really thinking with Teacher Tapp’s School Surveys – email me if you want to know more.
References
Friedman, D., Leeson, P., and Skarbek, D. (2019). Legal Systems Very Different from Ours.