Elgazette Logo newtrans  The newspaper for English language and international education

Comment: Performance pay or the highway?

melanie comment

Should language schools pay more money to their best teachers, and should student feedback be the deciding factor, asks Melanie Butler.

Personally I have always thought relating pay to performance was piffle.

When I was a book publisher, I was promised a bonus based on the number of books I published and their sales.

So when a fellow publisher had a particularly good year with a raft of new books hitting the classroom and soaring sales, she expected her annual bonus to go up.

It went down. Perplexed, she asked her accountant husband to explain the complex calculation on which the bonus was based.

He looked at it and laughed: ‘Using this algorithm they can give you whatever they like, and you can’t argue.’

Performance pay has arrived in language schools. All five of the UK language chains that submitted their pay results to the UK government gender pay gap exercise (see page 12) had bonus schemes.

But very few seemed to offer bonuses to their teachers. Surely teachers should also get performance related pay?

Common sense suggests that the answer should be yes. But the evidence is less than convincing.

According to the UK’s Education Endowment Foundation, ‘The results of rigorous evaluations, such as those with experimental trials or with well-controlled groups, suggest that the average impact of performance pay schemes has been just above zero.’

How much above zero?

A four-year-long randomised control trial conducted for the American federal government suggests that the answer is not much.

The difference between the end-of-year test results for the 65 schools that were randomly assigned to offer merit pay and the 66 that were not was between one and two percentage points, equivalent to three extra weeks of school a year.

The results were the same whether the schools based the performance scheme on test results for individual classes, the schools as a whole or on other measures such as classroom observation.

Worse still, an upward bump in test results occurred only in the first two years. After that, test results did not continue to improve.

Yet such bonus schemes do not come cheap. Since 2006, the US government has shelled out $1.8 billion to schools that introduce them.

In the private language school sector, paying bonuses for test results can be hard to implement.

For schools in the language travel business, it is well nigh impossible: the average learner coming to the UK, for example, stays just three weeks.

Instead, some schools are looking to base performance payment for teachers on another metric: student feedback.

Language schools are not the only bodies to take this approach. It is, in fact, fundamental to the British government’s Teaching Excellence Framework, which is designed not to increase pay for teachers but as a basis on which individual universities can raise their fees.

Once again, research evidence does not support it.

As Russ Mayne, who has looked at the studies in this field, explains on this page right, we know exactly which teachers get bonuses on this basis: attractive teachers who give good marks.

The good news here is that gender doesn’t matter. Which is not the case for most performance pay, even in UK EFL.

Take those five language school chains. Four of them paid higher bonuses to men than woman.

At Bellerbys the median average bonus was twice as much for men as for women.

Worse still, performance related pay rarely improves performance. Research published in the Harvard Business Review shows it works well for routine jobs like shelf-stacking or factory assembly lines. But in any job where learning and creativity are important, performance can actually get worse.

The researchers, Dan Cable and Freek Vermuelen, give three main reasons: First, the minute you introduce performance pay, you change motivation.

The intrinsic enjoyment of doing the work is replaced by the extrinsic reward of earning more money.

Second, all measurement systems are flawed. It is simply impossible to come up with a simple metric to measure a complex job.

Finally, as Cable and Vermuelen put it: ‘When a large proportion of a person’s pay is based on variable financial incentives, those people are more likely to cheat.’

That was certainly true in my days in publishing. Sales bonuses were paid on the number of books shipped out of the warehouse in any one year.

These were calculated in the January of the following year and bonuses paid out in February.

But books are sold on a sale or return basis.

So sales teams simply shipped out a mass of books before Christmas knowing most would be returned, unsold, three months later.

Just after the bonuses had been paid.