Like Russ Roberts, I’m skeptical of claims that, in general, a person’s welfare is reduced if he or she is offered more choices rather than fewer. More precisely, I’m skeptical of reading too much into experimental findings in which subjects seem to react negatively to increases, beyond some point, in the sizes of their choice sets.
But discard my (and your?) skepticism for the moment. The typical conclusion that scholars and pundits draw from behavioral-economics findings (such as the "excessive-choice effect," mentioned above) is that markets are weaker and less trustworthy than they would be if people acted more like textbook homo economicus and, therefore, the case for regulation by government is stronger than it would otherwise be.
Neither of these conclusions follow. In an interesting paper entitled "Less Choice is Better, Sometimes," Oklahoma State University economist Franklin Norwood builds a model in which consumers are harmed by having available to them too many product varieties among which to choose. But his model
also illustrates market incentives for recognizing and avoiding the excessive-choice effect. The use of advertising to raise industry costs can limit variety, leading to greater consumer and societal welfare. This is true even if advertising is associated with higher product prices. Perhaps the most effective method of providing the optimal number of varieties lies in the retailer. While any firm may produce and try to sell additional varieties, retailers have complete control over which varieties they sell. It is in the retailer’s incentive to provide the optimal number and type of varieties…. [T]here are many incentives and methods for markets to use to avoid the excessive-choice effect.
(Norwood’s article appears in the Journal of Agricultural & Food Industrial Organization, Vol. 4, 2006. Here’s the abstract.)
And in the current issue of Regulation, Harvard economist Ed Glaeser argues — powerfully, in my opinion — that "flaws in human cognition should make us more, not less, wary about trusting government decisionmaking."
Podcast RSS Feed
Full EconTalk Text
{ 18 comments }
That is a cool result. Too much variety is a bad thing? OK, the market will go meta on you and act to provide different 'numbers and types' of varieties so you can find one you like.
The "excessive-choice effect" elitism pisses me off. Is it any coincidence that it is promoted by people with an anti-market, anti-consumer bias?
At one time, the anti-capitalists criticized the free market because they claimed that it could not provide enough choices, and that it inevitably led to exploitive monopoly. Now, the anti-capitalists criticize the free market because it provides us with too many choices.
It takes time and effort to make a choice.
If you have n choices, that effort is roughly of (n log n) complexity.
Therefore, more choices equals significantly more time and effort to make a choice, which reduces consumer welfare *if* the average value does not rise accordingly. Each group of choices you add must raise the average value of the choices to avoid reducing consumer welfare.
Reducing this to individual choices, if the choice you are adding does not represent above-average value when compared to the existing choices, you do in fact reduce consumer welfare.
Now pay attention, NathanB:
USUALLY, THIS IS NOT A PROBLEM. Markets have a strong tendency to reject anything new which does not measure up to the status quo. Any difficulty generally arises when a great many complicated decisions are presented all at once, and the market at large is generally unqualified to evaluate them. This is usually the case when regulations are lifted and many previously blocked choices become valid, but not when no regulations exist – because new choices are generated sporadically, with more than enough time to be evaluated individually. The inadequate choices are then eliminated from the pool.
This has a positive-sum effect: if a new option is significantly better than others, an *older* option is commonly discarded. While n – and the (n log n) complexity of consumer decision – remains unchanged, the average value of this choice increases, increasing consumer welfare.
I would therefore propose that regulation is *dangerous*, because if it is eventually found to be unnecessary, its removal will almost necessarily reduce consumer welfare. Its presence, however, is intended to increase that welfare… and generally succeeds in the short term. In the long term, if the regulation stands and is just, the effect is minimal. The opposite is true of the damage caused by deregulation: early deregulation tends to cause minimal damage, while late deregulation tends to cause massive damage.
But I am not an economist, so I could be wrong.
n log n? You will pardon me if I'm not impressed by your injection of algorithmic terminology. Most people short circuit this problem by having brand loyalty and thus not devoting any time at all to their choice (Feynman's heuristic for choosing dessert also comes to mind). Time is devoted to evaluating choices only on a very intermittent basis, for example when buying something not bought frequently (cars or electronics) or when marketing has induced someone to try something new.
Agree with NathanB, plus there are plenty of "choice filters" such as net reviews, word of mouth, and consumer reports. I like choice.
n log n? I can now plan out my day better!
But, I am a little lost here…How does one go about messuring someones welfare at the moment of being presented a large choice of goods or services?
It is a brilliant strategy, you have to give them that. Nothing ensures having less choice than government interference in the market.
Perhaps we are seeing an evolution in modern liberal thought. First they pretended that government could improve the free market, or distribute wealth by force more ably than by trade. Now their arguments for intervention implicitly acknowledge government inefficiency.
All we need now is a modern day Hegel to come up with the synthesis – not only does government hinder free interaction among individuals, this free interaction is actually a good thing.
I'm not sure these lab scenarios faithfully reproduce the real world. In the real world, market segments form to address affinity groups, within which there are a smaller number of varieties.
E.g., I go to Whole Foods if what I want are a smaller set of expensive, chiefly "organic" products. I go to Kroger if I want mostly "mainstream" products at lower prices.
The internet is turning into a version of this phenomenon "on steriods", as affinity groups can be dynamically calculated, and personalization narrowed down to the individual.
Free market capitalism providing "too many choices" has got to be the last of my worries, really.
The whole "too many choices" idea suffers from the same problem as the "too many guns" idea. Who will choose who gets to have guns and who doesn't? Who will choose who gets to have choices and who doesn't? And yet, isn't that choosing itself? If too many choices are bad, then who could possibly take on the job of which choices should be eliminated?
Individuals only continue to evaluate more choices when they feel that the benefits of doing so outweigh the costs, so there is absolutely no way that offering more choices is going to harm a customer.
For example, presenting a customer with five million flavors of ice cream won't condemn him to months or years of weighing his myriad options. Sooner or later, (probably sooner) he's just going to pick something that looks good to him from the options he's seen so far. While some people might look at different options for longer than others, each and every one of those people will only keep looking as long as they feel that they benefit from doing so.
It is pretentious and dimwitted to substitute one's own value scale in place of someone else's and proceed to condemn the other person's behavior as being suboptimal. Whether or not Mr Norwood thinks that evaluating additional options is suboptimal is completely irrelevant to the satisfaction of those he is observing.
Ryan said, "…there is absolutely no way that offering more choices is going to harm a customer."
Except that it does. Too many choices can lead to anxiety, procrastination, and missed deadlines (examples: enrolling in medicare Part D, missing deadlines for camp enrollment because you can't decide). This is a problem for most people, but seems to be especially hard for the elderly, as well as people with ADHD. ADHD is a disregulation of inhibitory neurotransmitters which allow people to shut down excessive neurons that are firing. They have trouble making decisions because they are holding too many choices in their brain at one time. Everything is salient.
The practice of satisficing (purposefully limiting search, choosing the first one that is good enough) is a skill that must be taught. Few people do it automatically.
We should recognize that the bias towards completely libertarian free-markets is a bias — and I think it's a relatively good place to start from, as biases go. But psychology has much insight into actual human behavior that does not fit the rational-being model, and it is important to carefully examine the interplay between systematic cognitive biases and consumer behavior in markets. Opinion does not make for good policy.
All that psychobabble should be obviously irrelevant in light of the previous comments. The point is not that "excessive choice" harms SOME consumers SOME of the time, it is that those choices ultimately must be made by someone, and it is both economically inefficient and morally repulsive for that someone to be the state rather than each consumer.
@bbartlog:
"You will pardon me if I'm not impressed"
And you will pardon me if I ignore the rantings of someone who clearly didn't read the entire comment.
Of COURSE people commonly short-circuit the process, by eliminating the inadequate choices from consideration. I said that. Brand loyalty is simply an extreme case of this behavior, but it works the same way: the consumer identifies the lowest acceptable choice value, and then the highest (presumably basing the latter on budget). The consumer's choice is thus reduced to what he personally desires to handle, and becomes manageable.
In short, and to address Russell's concern, the market manages this automatically. You don't need to artificially restrict choices, because each individual consumer will restrict his own choices to make the process manageable.
The only problem is when you *do* artificially restrict choices, and then lift that restriction when pressured by the industry and the consumer market. The resulting glut of choices overwhelms the consumer's ability to choose, and they are compelly to set their filters not just suboptimally, but essentially *randomly* in the short term. This problem corrects itself over time.
This point has been already, perhaps stated slightly differently: if there is oversupply of choices in the market, then a free lunch is being ignored. That free lunch is profitably consumed by a reputable firm conducting a search and then making choices for its target audience. Given atomistic competition, there is selection in favor of firms which can best match product range and mix to preferences.
Caliban
"The only problem is when you *do* artificially restrict choices, and then lift that restriction when pressured by the industry and the consumer market. The resulting glut of choices overwhelms the consumer's ability to choose, and they are compelly to set their filters not just suboptimally, but essentially *randomly* in the short term. This problem corrects itself over time."
Why presume an over-correction when regulation is lifted? Your argument is based on search costs. But don't search costs lead to under-supply of variety in the short term, because people find it costly to seek out new products and therefore disproportionately favor incumbent products?
"The practice of satisficing (purposefully limiting search, choosing the first one that is good enough) is a skill that must be taught. Few people do it automatically."
You've got to be kidding me. Why, then, are video stores not flypaper for the elderly and people with ADD? If someone who hasn't been taught not to waste many hours evaluating their different options goes into a large video store, should we have a rescue team standing by to rush in and pick something for them?
When people go shopping for a car, do they look at every single car on the lot? Is it only their special training that allows them to avoid that fate? The elderly don't do this, and none of the people I know who have ADD (starting with myself) do it either. You're assigning an unusual pathology to the majority of the population in order to support a policy where some people limit the choices of others, supposedly for their own good. It's ridiculous.
"people commonly short-circuit the process, by eliminating the inadequate choices from consideration. I said that"
Not at all. You implied that this elimination had an increasing cost as the number of choices increased (via your 'n log n' claim – you believe perhaps that product choice is a sorting problem), whereas my claim is that the cost is fixed and negligible in many cases.
And I did read the rest of your post, but it was similarly full of personal theories stated as facts.
As for consumers being somehow overwhelmed by a sudden increase in choices, the brand-loyalty fallback still applies. The only examples of this kind of thing happening that I can think of are the culture shock that hit some soviet defectors when they came to the US in the 1980s (which may also inform your thinking), but they were not merely presented with a vastly increased number of choices, but deprived of their previous ones as well.