Merit pay and the candle problem

Let us pretend for one moment that many of the corporate education reforms being proposed offered more than just a metaphorical big stick with which to fire teachers more easily, but also a few carrots too in the form of extra money toward paying high performers as determined by their students test scores. Yes, yes, we know.

Let's go even further, and pretend that student test scores were the perfect means with which to judge the effectiveness of any teacher. What do we know of financial incentives? From Nature magazine.

here's a simple fact we've known since 1962: using money as a motivator makes us less capable at problem-solving. It actually makes us dumber.

In the 1940, an experiment was carried out, now referred to as the "Candle Problem". The experiment has the participant try to solve the problem of how to fix a lit candle on a wall in a way so the candle wax won't drip to the floor. The participant can only use (along with the candle) a book of matches and a box of thumbtacks.

Let's go back to that Nature article to explain the rest of the experiment, and it' counterintuitive results

The only answer that really works is this: 1.Dump the tacks out of the box, 2.Tack the box to the wall, 3.Light the candle and affix it atop the box as if it were a candle-holder. Incidentally, the problem was much easier to solve if the tacks weren't in the box at the beginning. When the tacks were in the box the participant saw it only as a tack-box, not something they could use to solve the problem. This phenomenon is called "Functional fixedness."

Sam Glucksberg added a fascinating twist to this finding in his 1962 paper, "Influence of strength of drive on functional fixedness and perceptual recognition." (Journal of Experimental Psychology 1962. Vol. 63, No. 1, 36-41). He studied the effect of financial incentives on solving the candle problem. To one group he offered no money. To the other group he offered an amount of money for solving the problem fast.

Remember, there are two candle problems. Let the "Simple Candle Problem" be the one where the tacks are outside the box -- no functional fixedness. The solution is straightforward. Here are the results for those who solved it:

Simple Candle Problem Mean Times :
WITHOUT a financial incentive : 4.99 min
WITH a financial incentive : 3.67 min
Nothing unexpected here. This is a classical incentivization effect anybody would intuitively expect.

Now, let "In-Box Candle Problem" refer to the original description where the tacks start off in the box.

In-Box Candle Problem Mean Times :
WITHOUT a financial incentive : 7:41 min
WITH a financial incentive : 11:08 min
How could this be? The financial incentive made people slower? It gets worse -- the slowness increases with the incentive. The higher the monetary reward, the worse the performance! This result has been repeated many times since the original experiment.

We've published a video on this phenomenon before, titled "As Teacher Merit Pay Spreads, One Noted Voice Cries, ‘It Doesn’t Work’", and an article from the Harvard Business Review, titled "Stop Tying Pay To Performance".

Here's another video - The surprising truth about what motivates us

Knowing all this begs the question, why are we going down the path of some of these corporate education reforms, when we have known for over half a century many of them are flawed concepts that have been demonstrated to fail time and time again?