Dave Hoffman over at Concurring Opinions asks: “Is Apple Exploiting Consumer Irrationality?” Dave is worried that consumers’ continuing iPod purchases may be irrational in the face of evidence that many iPod’s fail within their one year warranty period or shortly after, and that this strategy might explain Apple’s “growing market strength.” How likely are consumer biases to explain Apple’s success? In short, not very.
The claim seems to be that Apple’s product design decisions attempt to exploit consumers’ inability to realize that their iPod’s will break down sooner rather than later. Does this mean Apple could build a better iPod, but choose not to in order to increase its market share? Dave explains how the theory might work:
The optimism bias is among the most robust of the cognitive tics exposed by experimental behavioral law and economics literature. We consistently underestimate the likelihood of bad things happening to us. So, although Apple’s one-year warranty suggests a steep product failure curve at month 13, we discount that risk in our purchase decision.
Ahhh. The optimism bias. Two points here. First, I do not doubt that optimism bias is actually and meaningfully observed by some experiments, and is a real phenomenon on the individual level. As Jon Klick and Greg Mitchell point out, only a small number of subjects need exhibit a bias in order for it to achieve statistical significance. That said, the power of even the most robust of these findings, such as the “endowment effect,” have been (at the very least) called into question by recent research. I am quite skeptical that the optimism bias is a leading candidate for explaining Apple’s success.
Second, the question is, and always has been, what happens to individual biases in markets? Do markets serve as a “debiasing” force? Or do they exacerbate individual biases? If the second is correct, what are the efficiency costs of the bias? I understand that Dave was not addressing this issue in his post, but if it is true that these biases are allowing Apple to dominate the market, well … I have read enough behavioral economics-based paternalistic regulatory proposals to know what is coming next: how do we
regulate fix it?
So, before we turn to that question, let’s sketch out some of the economics here.
As a building block, let’s think about how competitive suppliers of iPod-like machines would respond to the presence of such biases. To make it exciting, make some assumptions that strengthen the impact of the biases: (1) all consumers exhibit them; (2) education cannot ameliorate their effects (“once a bias, always a bias”); (3) sellers are aware of (1) and (2); and (4) consumers who buy one type of machine are “locked in” for repeat purchases.
Ok, so what happens in the competitive market for iPod-type machines when consumers do not have the ability to correctly “price” the durability of the machine (or the short length of the warranty)? Because each consumer is biased in this manner, the initial sale is extremely valuable to the seller since it increases the probability of future sales. But in a zero profit equilibrium, sellers (aware of these future, bias-induced sales) will compete away the supra-competitive return by offering consumers favorable terms. For example, lowering the price. Completely biased consumers still receive the benefit of competition because sellers compete for the right to sell to them and appropriate the profits associated with the bias. This is good for consumers.
But, the allegation here seems to be that this is not a competitive market. Rather, Apple is presumed to have power over the pricing of iPod-type machines. The economics change now because Apple does not live in a zero profit equilibrium — so NOW wouldn’t this theory of anticompetitive harm be viable? I dont think so. If Apple is a monopolist, it strikes me as implausible that they would collect their monopoly return through repeat purchases induced by having machines disappoint consumers by breaking down early.
Perhaps Apple is waiting patiently for the day when it will use its market dominance to abuse consumers? As one commenter to Dave’s post expresses this concern:
“who knows when they might suddenly decide to leverage that dominance into an effort to force all iPod users to buy Macs? “Just buy a Mac desktop” was suggested to me more than once by the guys at the ironically named “Genius Bar” at my local Apple store.”
Wait. Even a monopolist is allowed to allow its salespeople to suggest that you buy their products. In the meantime, Apple is offering substantial benefits to consumers by improving its product and developing complements. If we are truly concerned about the risk of future anticompetitive harm, wouldn’t it be wise to allow consumers to receive those benefits until the risk materializes? We are not talking about the costly “unscrambling” of a merger that turns out to be anticompetitive.
Of course, if consumer irrationality was the culprit, one might wonder why competitors do not simply shed light on the short product life of the iPod. I also wonder how one might map irrational consumer behavior into meaningful policy predictions. For example, do consumers with optimism bias sanction iPod for supplying products they feel are defective? Or does their optimism persist? Dave notes that:
This optimism is no doubt enhanced by Apple’s careful packaging, which makes it look like they’ve taken a swiss-like level of care in their manufacturing process, and iPod’s high-price, which suggests quality . . ..
This analysis suggests that so long as Apple retains its brand – expensive, low-defect, attention to detail – it will continue to convince consumers to buy products with lower-than-expected lives.
If only market success (and a monopoly rate of return!) was as easy as a fancy gadget in shiny packaging that breaks down early and disappoints consumers. I am in the wrong profession. Interestingly, Dave notes he will continue to buy another iPod despite his awareness of consumer optimism bias. Perhaps Apple’s success is not driven by optimism bias after all?
Doesn’t Apple have two revenue streams, one from ongoing song purchases and one from equipment purchases? Which is the more substantial one?
When the hardware breaks, the customer is confronted with the whole system decision again. The ideal case for Apple might well be that each iPod lasts forever, as long as people buy songs (which in turn really cements Apple’s market position).
What measure of “market share” is assumed? Number of units sold or number of active/intact units in circulation? Isn’t the first a bit dysfunctional if it doesn’t correlate closely with the second?
I donâ€™t have an iPod and I canâ€™t follow the intricacies of the dialog above, nor do I want to speculate about the efficacy of regulatory intervention, but it does seem the generalization that marketing (which traditionally includes product design) is keen to the kinds of considerations that behavioral economists focus on is correct. But for bounded rationality and other biases, it is hard to imagine that super bowl advertisements would be very fun at all.
Josh: Great post!
Given the type of good Apple is providing, I’d want to see a lot more empirical data before I bought into Hoffman’s analysis. One thing he doesn’t contemplate: iPods have very short product cycles in any event. For instance, I’ve joined the iPod brigade with a Nano, and I’ll probably keep it. But a year from now, who knows what Apple will have come out with? If it broke, I might upgrade, but I might also change brands.
It’s worth noting that I changed to the iPod after one of the competitor’s products (a Creative Zen) broke on me after 30 days of use. Nocera seems to be assuming that there is a higher-quality competitor out there.
Hallelujah. I now understand why American cars are so awful. Thatâ€™s because the Big Three are exploiting the optimism bias. GM could have made fantastic cars, but chooses to makes Chevies bad on purpose, so that consumers (who are of course locked into buying Chevies forever) would buy one defective Chevy after another. After all, GM is a monopolist â€“ it has a monopoly on manufacturing of Chevies!
Matt, I dont think that you are showing ignorance at all. The key trade-off here is exactly the one that you have identified: the cost to Apple of engaging in this type of strategy against the potential for increased duration of monopoly return. You suggest that switching costs associated with building up the library after the initial year will mitigate any substitution towards rivals which will hurt Apple. Maybe so. But I think these effects should be relatively small to the economic forces driving Apple to improve its product relative to the world of potential substitutes that may dethrone the iPod, and to earn the profits in the myriad of complementary markets. It makes sense to me that these economic forces are far more important to the competitive dynamic than these sorts of timing devices. Further, I am not sure how important these switching costs are from a welfare perspective, or how one would reasonably go about “solving” the presence of switching costs stemming from consumer preferences. I guess I have already hinted enough that I do not really think this is much of a problem at all.
It should be noted (and not in direct response to Matt) that ex ante “network” or “systems comeptition,” i.e. competition for the field, is also an important part of the competitive process and creates further benefits for consumers. Competition for the opportunity to profit in complementary markets (or with repeat purchases resulting from switching costs) increases consumer welfare as manufacturers fight to get broader distribution. Geoff alludes to this point in his comment. Frequently, the gut reaction in antitrust analysis is that ex post competition (on the retail shelves for consumers) is more likely to increase consumer welfare. This reaction, in my view, is fundamentally incorrect.
As a non-iPod owner, I may be showing some ignorance here, but why do people buy iPods and not a cheaper substitute? I’ve assumed it’s because Apple is the only game in town, as far as song selection, ease of use, and sound quality go. It’s why college students gobbled up Macs in the late 1980s — they was nothing else really close to them. At some point, however, there will be substitutes. Apple’s goal is to stretch out their market dominance as long as possible while extracting the highest amount of profits that they can.
You assume that they wouldn’t want a machine that breaks down in one year, rather than four, because it would hurt their market share. And yes, over time it would. But how long would it take for consumers to realize that their iPods break down? At least a year. And by then, they’re already stacked with songs. Word may spread eventually, but it will be slowest in getting to the folks who buy their iPod later in the game — i.e., the less cutting-edge folks. So if Apple times it correctly, then the market reaction to the short-lived iPods doesn’t happen for three or so years — perhaps right about the time when there are real competitors on the technology. And then, voila! — Apple can announce a new, improved, longer-lasting iPod.
Look, it is fairly paranoid to think that Apple is rigging the iPod to wear out more quickly — the more logical inference is that they’re stuck with the technology they have. And I agree — if they can keep their monopoly longer by producing the best product, then it probably makes sense for them to do it. But it’s the old tipping point question. Once consumers have invested, monetarily and psychically, in their iPods, how much will it take to drive them away?
Matt, I think you only responded to half of the point that I breezed over too quickly. Of course, if Apple is a monopolist, it is plausible that they would appropriate their monopoly return through the price of the product — that is the most reasonable way to do it. The claim I responded to was that Apple was extracting its short-term monopoly profits by inducing repeat purchases as a result of its shoddy product design, thereby exploiting consumers’ optimism bias.
Here is why I find that claim implausible. And quite frankly, it is even MORE implausible without the optimism bias:
If you are a monopolist, looking to get your monopoly profits while you can, the strategy of selling products with poor design to induce repeat purchases has a very significant cost. Specifically, of losing your monopoly position without ever obtaining your monopoly rent!
The reason that you are a “short-term” monopolist is because there are not yet sufficiently close substitutes to your product to discipline your pricing. The longer you sustain your monopoly position, of course, the better off you are. Bad warranties and poor product design on the hopes that you induce future sales seem like an odd way to extract monopoly rents in a world where you may lose your position at any moment as a result of the innovative efforts of entrants and competitors. Even a one year wait in the iPod world is not trivial in terms of the potential for technological change.
I think Josh breezes over this point a little too quickly:
“If Apple is a monopolist, it strikes me as implausible that they would collect their monopoly return through repeat purchases induced by having machines disappoint consumers by breaking down early.”
Why is it implausible? If you have a short term monopoly, and you’re looking to extract your monopoly rents while you can, why not raise the price of the machines? And is it easier to raise the price by raising the actual price or by shortening their shelf life? For example, do you think they would sell more iPods if they cost $1000 and lasted four years, or cost $250 and lasted one year?
So you don’t have to get to the optimism bias at all. But that’s far from saying they are intentionally shortening the iPod’s longevity. That’s quite another leap.
Josh rightfully dismisses the “tie” as evidence that competition is lacking. iTunes music — part of the tie you mention — is part of Apple’s product. Apple offers not just the iPod, but the iPod plus tons of music, maximum trendiness, beaucoup accessories, etc., etc. And people buy the iPod for all of these reasons (and reject the myriad network contenders that donâ€™t quite stack up . . . at least for now). The music is as much a part of the product as the device itself, at least for most customers. So how can it be that Apple’s product superiority is a quirky, worrisome reason (a “tie”) why competitors can’t find purchase? Isn’t that just competition, full stop? The worry that eventually Apple will flip the mysterious switch back at HQ that renders all of our iPods (and iTunes music) useless unless we buy a Mac is fanciful. What would be the theory? That there is no price too large for iTunes music collectors to pay to maintain their libraries? I have 3 quasi-words for you: LPs, CDs and tapes. Inter-network competition is real competition; it isnâ€™t only the intra-network kind that matters. Besides, by the time Apple does flip the switch, we may well have gotten more than our money’s worth, anyway.
You might argue that optimism bias creates a window for Apple to abuse its customers (perhaps not large enough for “buy a new Mac or lose your music,” but maybe big enough for some abuse â€“ like planned obsolescence or shoddy construction). Perhaps, but here I have to ask â€“ how reliable is the evidence of a bias like this? Doesnâ€™t it depend entirely on discount rates? You note that people are more optimistic than the models predict. How do we measure that? I have a pretty high discount rate â€“ am I “irrational” or “overly optimistic”? Just curious about the quality of the claims that weâ€™re less rational than we think we are.
Nice post, Josh. I confess I too am skeptical of Nocera’s argument that consumers are ill-used by Apple’s warranty program. The optimism bias was my attempt to make the best case possible for his argument.
I wonder why, though, you so quickly dismiss the commentator’s claim that the reason competitors can’t find purchase is the tie. A large part of the reason that I’m going to buy more ipods is that I’ve invested heavily in itunes music. (That said, I too think we ought to wait until real harm materializes before “fixing” anything through regulation or litigation. I posted on conglomerate a while back about the rise in Mac sales as a part of the iPOD halo effect – you will see that I’m a part of whatever trend line exists.)
I agree with you that simply saying “optimism bias! (oh my)” isn’t a sanction for paternalistic interventions into a market. But the bias does help to explain why consumers do not process information about product life in the ways that the rational actor model would predict. That is, even if competitors were to run ads saying “IPods Die Like Mayflies,” the bias suggests that at least some consumers would not process the message and change their demand.