Error propagation and extinction

One of my minor neuroses is an aversion to propagating errors of fact or logic. Indeed, I have to apply teeth to tongue at times when witnessing others propagating error. Managing this quirk productively is an important part of pedagogy, as experienced MBA instructors will immediately recognize. (Note that you will have many more opportunities to correct errors than to answer questions, because part of not understanding something is often not realizing it.)

Knowing when to pull the trigger on a correction to an error is the most subtle aspect. The first-best solution is another student immediately chiming in with an on-point critique, but that happens rarely. A lightly guided discussion that eventually corrects the error is next best, but there are practical challenges here as well, since a) limited class time may be available to deal with the topic, b) it can become aggravating for the students to play “guess what the professor is thinking,”, and c) the longer the uncorrected statement lies there the more likely that students will internalize the error and repeatedly spout it back in future classes, on exams, etc.

Assuming one has let the error go uncorrected as long as seems prudent and decided to directly intervene, it’s still often a challenge to a) precisely recognize the nature of an error and b) quickly come up with a concise, memorable, and understandable correction that will persistently displace the erroneous idea from the audience’s minds. Of course, experience helps, because errors tend to fall into repetitive patterns, allowing you to build up an internal database of diagnoses and appropriate responses. Here are some classic sallies with proposed responses below the fold. Suggested improvements to these responses (as well as additional examples of “favorite errors”) are welcomed in the comments.

1. “The company has a cost advantage because it makes more products and there are economies of scope in this industry.”

2. “The company has a cost advantage because it’s more vertically integrated (and Porter says that reduces costs).”

3. “The company has a cost advantage because it outsources more activities.”

4. “There are strong entry barriers because small companies can’t afford to pay the capital costs to operate in the industry.”

5. “The company needs to give better deals to its loyal customers.”

6. “The big growth in this industry comes from this new segment X, so the company should focus its resources on penetrating X.”

1. “In general, increasing product variety (holding total sales constant) increases unit costs, even with economies of scope. Often, selling more varieties increases a firm’s total sales and that may beneficially exploit scale economies or increase capacity utilization, but a firm can rarely reduce operating cost per unit solely by increasing the complexity of its output mix.” (Optional algebra explanation: Economies of scope says only that F1 + F2 > F12, not in general F1 > F12 or F2 > F12, where F1 = fixed cost of producing only product 1, F2 = same for 2, and F12 = fixed cost of producing both varieties together.)

2. “Is that a general rule? It couldn’t be, or else all firms would make their own electricity and paper clips. And how do you explain the long-term outsourcing trend?”
“There are specific conditions under which we would expect vertical integration to reduce cost. Ask yourself what efficiencies the two firms can effect by being jointly owned that they could not while separately owned.”

3. “The logical implication of that argument is that vertically integrated companies are either at an overall disadvantage or else generate higher quality and hence higher willingness to pay. But are all vertically integrated firms at a cost disadvantage? Why do paper companies mostly make their own pulp and aluminum smelters mostly make their own alumina, given that these are largely commodity products with small quality differences?”
“There are specific conditions under which we would expect vertical integration to reduce cost. Ask yourself what efficiencies the two firms can effect by being jointly owned that they could not while separately owned.”

4. “But what about big companies that might enter? Remember that entry barriers are about the most-qualified potential entrants, not the vast army of no-hopers.”
“Is the problem really a lack of financing? If you personally had the money to finance it, would you make the investment to enter this industry? If not, the entry barrier is not likely to be financing constraints but rather something causing lack of expected profitability.”

5. “Depending on what you mean by loyal, it might make more sense to soak the loyal customers and give better deals to the customers where the price concession is more likely to affect their purchase decision (i.e. the “disloyal” customers).”

6.  “The ‘so’ in this argument needs to be bolstered. You have to consider how profitable X has been for the current players, whether the company would be equally competent as the current players in X, whether the company’s efforts in X might cannibalize more-profitable sales in other segments (that otherwise would not be lost), and whether whatever resource constrains the company’s growth would best be allocated to X or to other segments (including possibly its current ones).”


2 Comments on “Error propagation and extinction”

  1. I, too, hate error propagation — but I don’t think from a position of authority is the best way to make the point. You’re trying to ANSWER them. Instead, I’d recommend that you QUESTION them further until the original logic falls apart. Destroying their (and their classmates’) belief in their original answer is your goal.

    1. “The company has a cost advantage because it makes more products and there are economies of scope in this industry.”
    Q1: “How does the company save money by producing both Twinkies and Cupcakes?”
    Followed by “How far can we stretch this?” as in your vertical integration answer.

    2. “The company has a cost advantage because it’s more vertically integrated (and Porter says that reduces costs).”
    Q2: “Where are we saving money by making our own inputs, and how much are we saving?”

    3. “The company has a cost advantage because it outsources more activities.”
    Q3: “Tell me some of these activities that we’re bad at, but our suppliers are good at.”

    4. “There are strong entry barriers because small companies can’t afford to pay the capital costs to operate in the industry.”
    Q4: “Are you referring to the initial capital investments, or the ongoing costs to service their debt or provide returns to their shareholders?”

    5. “The company needs to give better deals to its loyal customers.”
    Q5: “Why should those loyal customers get the best prices?”
    Followed by “What’s the argument for giving the best prices to people who aren’t currently buying our stuff at all?”

    6. “The big growth in this industry comes from this new segment X, so the company should focus its resources on penetrating X.”
    Q6: “What will happen if we completely ignore segment X?” and/or “Can we make money in segment X?”

    I sometimes use the rhetorical device of restating the student’s point and ending my restatement with “…right?” Students quickly learn that anything I end with “…right?” is dead wrong.

  2. stevepostrel says:

    That’s a good point. About half of my suggested answers have questions embedded, so I’m certainly not hostile to that approach. It has some clear advantages, like getting the students to keep thinking about the issue and possibly making the aha! more memorable when somebody gets it. A good poser like that stands a chance of displacing the erroneous idea from the student’s mind, a necessary step before they can grasp a better concept for dealing with the issue.

    But in the end, I also would like a concise explanation of a) why the one way of thinking is confused and b) what the better way is. Many students internalize through memorization before rumination, so having a brief explanation seems like good follow-through. Also, I fantasize that when the students are later faced with the same error from their colleagues they’ll be able to deploy the explanation they learned in my class and hence contribute to anti-error propagation.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s