Posted August 30th, 2012 by bbsadmin & filed under Education.

I usually don’t focus on education, but our future will depend on how well our kids understand the world so I borrowed the following from Free Exchange because I thought it was excellent

HOW do you improve education? To economists the answer is simple. Pay teachers for performance: if the pupils get good test results, give the teacher a bonus. Attempts to incentivise US teachers to bump up grades have generally proven ineffective, however. The solution, according to a recent research paper finds, is to hand teachers a large sum in advance and dock their pay if students flunk their exams. This gets results.

The authors of the paper divided Chicago teachers into two groups: a “loss” group and a “gain” group. They paid “loss” teachers a bonus of $4,000 at the start of term. If exam results were below average, they took away up to $4,000, depending on performance. If results were above average, teachers could earn an additional sum of up to $4,000. “Gain” teachers were simply paid a bonus of up to $8,000.

The same performance entailed the same bonus in each group. Yet the “loss” group lifted standards while the “gain” group did not. “Loss” teachers raised maths test scores by 0.2 to 0.4 standard deviations. These effects are comparable to lowering class size by more than a third. “Gain” teachers raised maths test scores by small and statistically insignificant amounts.

The authors put the difference down to a phenomenon economists call loss aversion. Humans tend to fear a loss more strongly than they desire a gain of the same value. Here, a contract that offered a gain provided little impetus to work harder. A contract framed in terms of losing $4,000 produced a greater urgency to improve.

Should this be translated into state policy? There will probably to be resistance to such a scheme. Teachers’ unions might see the policy as a cover for wage cuts. They might disagree with a system that piles more pressure on teachers to improve. The idea will likely be caricatured by the notion of mischievous kids saying to their teachers, “Sorry Miss, you lost your bonus. I got an F.”

Practical issues aside, there are reasons to be cautious. First, there is a risk of cheating. Tie teachers’ pay to kids’ grades, and there is a greater spur to fiddle the test scores. Brian Jacob and Steven Levitt (co-author of “Freakonomics”) looked at incentive regimes in Chicago schools in 2003 and found that a “minor” change in incentives caused a large spike in cheating.

Second, teachers could urge their weaker pupils (most likely to be the poorer ones) not to take the tests. This would raise average performance leave slower students behind. Third, there would be greater pressure to stick solely and rigidly to the test syllabus. Teachers hoping to imbue kids with a love of the discipline might dislike a more test-focused system.

The authors found no evidence of the first two things in their new study. But this is not to say school incentive schemes as a whole cause no behavioural change. Papers by David Figlio find that “incentivised” teachers come up with ingenious ways to boost test scores. Menus are changed nearer test day, to boost calorie content. Unruly pupils are suspended prior to exam season. Marginal pupils are reclassified as disabled. There are numerous ways to game the system, many of which cannot be properly policed or checked.

In the long term, the change may be more fundamental than economists realise. You are transferring from a system where the agents are (to a degree) public-spirited individuals to one that motivates agents to be self-interested. Cheating and deviousness apart, is this what we want? Do you end up attracting a wholly different kind of teacher to the job, one less interested in the subject itself? Do you put off the Dead Poets Society teachers in the process? One can’t be sure. But this would be hard to measure.