On Choice

September 11th, 2008 admin Posted in Experiences, Philosophy |

My friend, Sandeep, sent me a link to a very interesting talk given by Dr. Barry Shwartz, a sociologist, who observes and persuasively argues that excessive choice is bad.

I agree with this observation. More generally, this observation applies to any kind of decision making. We make many decisions in life - in everyday life and in the grander scale of life. In everyday life, we decide on things like which vacuum cleaner to buy, or, which hair conditioner to buy, both of which were decisions I had to make last week. In the grander scale of life we need to make decisions such as who to marry, or, which profession to work towards.

To make a decision we weigh pros and cons across several dimensions and finally settle on a decision. If we do a good job at making the decision, we have considered all the dimensions and all the pros and cons along those dimensions. We are happy that we have come to a global optimum across the search space of the solution, and we can pat ourselves on the back for it. As the number of dimensions grow, however, the search space explodes in size. The resources and time to do this optimization become overwhelming. We are forced to work with a smaller search space that we can handle within the time we give ourselves to make the decision. Say we have an hour to kill and feel like watching come TV. We sit in front of the box, pick up the remote and flip through channels. We want to decide which channel to watch. We do not want to spend that hour trying to optimize that choice. To make a decision in some reasonable time, we need to make more decisions first. We need to decide which dimensions of choice we want to ignore. This is choice-pruning. For example, we might not care for a show if it is not in Hi-Def. Then the choice-space shrinks by an order of magnitude. However, once we do shrink the choice-space, we have to settle for something that may be a sub-optimal solution compared to the global optimum. And that can take away from the satisfaction you draw from your, potentially sub-optimal, choice.

For some of these decisions there is no way to know the optimality of the solution. For example, once you marry a person, you better make appropriate adjustments and make happiness out of it. There is no point comparing the decision to anything - what if I had married someone else, what if I had taken up that other job offer 10 years ago? In some ways not knowing the optimal solution is a boon. You can rest assured that there is no such thing as an optimal solution in that case. The decision you make is the only decision that matters. No one can prove to you that a different choice would have definitely been better. It is calming to know that your decision is beyond judgment. Still, too much choice can be paralyzing in this case also. The only saving grace is that almost any decision is really a pretty good decision.

For some other decisions however, there is a direct measurable impact. These are the decision that can haunt. The stock you choose to invest your money in may tank, while that other option you were considering just as fervently, does really well. The vacuum cleaner you chose sucks, while the one that your friend bought for just a little bit more money, sucks better. And it is for these decisions that too much choice can not only be paralyzing, but also be humiliating. You blame yourself for the sub-optimality of your choice; after all, you have proof to justify that blame.

I routinely find myself presented with this overdosage of choice. And being mathematically and engineeringly inclined, I tend to at least give choice-pruning and optimization a fair shot. As I wrote to Sandeep in my response email:

“Being an engineer I tend to compare the available choices across many dimensions, and the search space grows multiplicatively. I was looking for a hair conditioner yesterday, the dimensions were - ingredients (should not have any obvious bad stuff), the company (should be something I have heard of), the quantity (should not be measured in gallons), the price (should be reasonable), delivery mechanism (spray vs cream), application time (dry hair or wet hair - further subdivided into needs rinsing after application or not) etc. Comparing all this across the 30 brands in the store left me tired.”

In fact, excessive choice wastes time and energy, and therefore, wastes money. Just by spending 20 minutes to decide on a hair conditioner, I am sure I bumped up the price of the produce by a few dollars. Time is money, and this was time I could have better spent elsewhere.

And the solution to this problem of excessive choice is not communism, or even reducing choice, necessarily. Availability of choice is not the root problem. The problem is the excessive demand excessive choice places on the decision-maker. The decision-making has to continue to be streamlined. The choice has to be presented to the chooser in a structured, standardized, unbiased way. This is really a service, which I like to call “choice pruning”. It can be an industry in its own right. An example of this is epinions.com, where people’s past experiences with their decision-making about buying a product are collected, analyzed and presented to the customer to help speed up his or her decision making process. The opinions are not strictly standardized, and I can spend days reading through the reviews there, as I recently did when trying to decide on a vacuum cleaner; but still, it does help. At least you know that with hundreds of respondents, there is a chance that any biased views and any person-to-person variation in the interpretation of the measurement scale are evened out.

By spending time on what we are choosing, we affect what we are choosing. If I spend 5 years to decide which stock to invest money in, I have already lost more money than a sub-optimal decision could have cost me. In other words, as the time to make a choice increases, by the time you finally choose, the choices available to you might end up being different than the choices you optimized for! Thus anything to speed up the selection, anything to assist with the optimization, anything to reduce future repenting, is a much-needed solution. For me, rigorous, instinctive, choice-pruning, indifference to the actual choice made, and a poor memory, help make this process faster, reasonably optimal, and guilt free.