The Law of Nice Numbers: Incidence of Anchoring

I find numbers to be interesting, both in an abstract and real sense—take the robust and widely observed behavioral phenomenon, anchoring. Even if you haven’t heard about it, you have certainly succumbed to this bias, and there is a way to take advantage of it.

But first, what’s anchoring? Put simply, it’s the tendency to use irrelevant information as a basis for answering (in most cases numerical) questions. Consider the classical example:

In Kahneman and Tversky’s “Judgment under Uncertainty: Heuristics and Biases” (1974), project participants were spun a wheel of fortune rigged to land on either 10 or 65, then were asked: What’s the percentage of the United Nations that are African nations?

Obviously, neither 10 nor 65 have anything to do with the real value, but interestingly, answers differed significantly depending on where the wheel stopped: Participants whose wheel landed on 10 made significantly lower guesses than participants whose wheel landed on 65.

And yet, after decades of research and studies, we’re still not sure how the mechanism itself works. Here are three prevailing paradigms:

  • According to Kahneman and Tversky, people use the initial value (the anchor) that is adjusted to find a final answer (the initial value may be suggested exogenously or be a result of computation), however, this adjustment is typically insufficient. They pinned the term “Anchoring and Adjustment.”
  • Currently though, their explanation is challenged by the Selective Accessibility Model based on Confirmatory Hypothesis Testing (Furnham and Boo, Chapman and Johnson). According to this view, we access similar information to the anchor (in our memory, of course!) and use it to check the hypothesis, whether the anchor could actually be the true value.
  • Alternatively,  Attitude Change suggests that anchors themselves might indirectly influence information processing mechanisms introducing biased judgments and thus anchoring. Thus suggesting U-shaped response, meaning that with more extreme (less probable) anchors, the anchoring might be less evident.

But what’s fascinating is the ubiquitous nature of anchoring. Let me try you, dear reader. Tell me:

1. How much should the engagement ring cost? Give me the number of months’ worth of salary, please.
2. Which charity would you be more likely to donate $5 to, according to their supporter brackets?

i.      Charity 1 : $5, $10, $15, $20

ii.     Charity 2 : $5, $50, $500, $5000

3. You go to a grocery to get a single can of cola. Good news! There’s a 10% discount for it limited to 5 units per person. How many cans would you buy?

It’s not the exact values that are important, but the irrelevant ones. You are very likely to base your decisions on them in this type of situations.

In fact I would argue it’s the anchor that might have drowned the rational agent described in economic theory, which is inscribed in our society and even customs (the engagement ring!). Thus our brain uses shortcuts or biases to solve problems using less effort. They’re lazy, as Kahneman would put it.

You’d think those who use numbers often, for example investors, would be less susceptible to this type of bias, right? Think again. The opposite is more likely to be true.

Anchors almost always work as benchmarks to manage losses, but unfortunately work against investors. For example, if the price of an acquired asset drops, the starting price acts as a level that investor wants to be reached again before selling the asset. Even worse, some investors (if not the majority) use previous high of an asset that has declined in value as an anchor to assess its cheapness.

An even more interesting occurrence is the Law of Nice Numbers. I have to admit that it was not me who came up with this phrase, but my friend. It captures the idea nicely. Ever wondered how nice, round numbers work as a support or resistance level for assets? It’s quite self-evident, consider 15k for DIJA or 600$ for Apple’s stock. Even better, consider the mystical 666. Did investors get superstitious over the crashing market during GFC?

Figure 1: S&P’s interesting resistance point

1.png (1073×831)

Source: Free Stock Charts

So what should you do about it? There is couple of things one must keep in mind:

  1. Going to a job interview? Uncomfortable with the salary question? Give them a big number and don’t be shy, anchoring works against them in this case.
  2. Selling something? Set a big starting price.
  3. Buying something? Counteract with an aggressive offer, to re-anchor the negotiations.

Generally, if you want people to be lazy and not to think too much, give realistic, nice and round anchors. If for some reason you want them to engage – do the opposite.

And always… always use your own method of deduction. The given numbers are most likely irrelevant but harmful. To be concise – think!

One Response to "The Law of Nice Numbers: Incidence of Anchoring"

  1. David Stackpole   August 6, 2013 at 10:59 am

    What an excellent example of human bias, all to prevalent in business and investment finance.