Experience Versus Information

I am reading David McCullough‘s The Great Bridge, about the building of the Brooklyn Bridge. (Yes, I like McCullough very much.) In a passage about the fundamental differences between Brooklyn and New York (i.e., Manhattan), McCullough writes:

People then were still inclined to form opinions more from experience than information and it was the experience of most Brooklyn people that between their city and the other one, there was no comparison.

The time he is writing about is the late 1860’s. The observation that fascinates me is that “people then were still inclined to form opinions more from experience than information.” This strikes me as both a) true and b) profound, though I am willing to be argued out of either one. It also makes me wonder further:

1. In the modern age, do we primarily treat information as a substitute for experience?

2. If so, do the benefits outweigh the losses?

3. If McCullough is right, doesn’t it make bad information that much more dangerous?

4. If McCullough is right, might we approach a point where information distrust and overload encourage people to return to experiential wisdom?


I would love to know what computer programmers think; they start with knowledge and earn experience along the way.

Knowledge has spillover effects (positive extenalitie) and I would think is used more frequently becasue of increasing returns to scale. That is one of the reasons that incubators and agglomeration economies work.


Interesting question. There are two instances in which I'll try to test it: 1) What is my opinion about an issue in which I have no direct experience; and 2) If asked about an issue in which I have direct experience, but also information that contradicts it, what opinion do I then hold?

In the first case, if you don't have direct experience, you easily rely on information you've collected from others with direct experience or other secondary sources. If that is limited and you still want to form an opinion, you make something up based on other experiences that may or may not be similar.

But, in the second case, I think people are more apt to opine based on their own experience even if all the information that they have gathered from other sources contradicts this. An easy example for me is answering the question, "Is Detroit a violent city?" Statistically, yes. I know this based on reliable sources. But, having lived there, my opinion is that it is no more dangerous (this is not true) than Los Angeles, NY, Chicago, DC (all places that I have lived as well).

I think the danger is that many people don't consider the contradicting information that exists; and, in fact, take great pains to dispute and dismiss it. This is a unique type of arrogance that greatly impedes intellectual debate and exacerbates generalities.



I would stay the vast majority of people still make decisions based on experience. A certain class/type of people now base their decision on information, though, yes. Especially with the increased availability of data on the internet. However, as we can see by a proliferation of (very popular) but crummy sources of information, such as Fox News, many of that information does indeed prove bad. As Farhad Manjoo ("True Enough") notes, there is an overabundance of bad info out there creating multiple realities of what is "fact" and what is "fiction."

One of my complaints about many economists is that they live in a world on paper where information is perfect and nobody makes experience-based decisions. Sunstein and Thaler ("Nudge") note that this is a problem and that we're Humans, not Econs.

Solution: education. Fact: on the whole, we do it rather poorly.

joe.shuren, bouvet island

How would you design some experiments to settle this question? Define the problem (and in that is implicit the possible solutions). Brainstorm for alternative hypotheses. Identify the scientific researchers who are interested in the question. Figure out a way to raise some money for a committee to supervise the work. Make some controlled, double-blind experiments using computers with different cultures and subjects. Publish the results and try to confirm them. Then write about it in the New York Times. In order to be of value, you would have to translate the scientists' rhetoric into narrative anecdotes, video stories that make sense to the viewers' experiences and add not only information but actually change behavior.

Information overload is not only too much information, but information that is not at the right time, and that is not organized in a pattern that makes sense. Who decides the answers, and how?



1) If by "we" you mean "everyone, except those who think critically (e.g., Freakonomics readers)", then YES.

2) If you do not think critically, then YES--you know more. If you do think critically, then NO.

3) If the "bad" information adheres to what you already believe, then NO--see, you were right. If you think critically about the information you come across, then still NO--you will reject the "bad" information, and gain insight about the type of person who accepts it.

4) If by "people" you mean "everyone, except those who think critically (e.g., Freakonomics readers)", then NO--what's to be distrusted when there's plenty of information that supports what you already believe?


As a golfer, I can say from experience that opinions formed from information about golf are definitely much different than opinions formed from experience with golf.

However, the basic concept really needs to be reworded to be more specific. The question is this: are people's mental models about X are created more from experience with X than information about X?

When you look at it this way, it becomes a lot easier to answer the four questions.

1. For a lot of thing, yes. Information = encoded experience of others. You can make your model better by reading the experiences of others...as long as your model isn't too far off.

In fact, large parts of society assume that you have no direct experience in the area that you're supposed to understand...like government financing.

2. Yes, of course. Your model can get better without having to take the time and energy to actually do it. There are limits, of course, but the more experience you have, the easier it is to update your mental model using someone else's information.

3. Bad information is always dangerous.

4. Well, that's impractical, and really the question means nothing. Ideally a mix of practical experience and abstract information is what you want.

If you really want to think about the problem clearly, use sex.

Is information about sex different than experience with sex?

Is bad information about sex much more dangerous?

etc etc


Remembering gramma

Some of you may remember the expession shmata. Its a jewish expression that means something like piece of clothing. It could be an old coat that you are wearing, a pocket book. Usually, it refers to clothing that's tattered and old and ready to throw out. Well, the point is there is nothing that the word data could encapsulate that suggests a shmata. Think about it.

Michael F. Martin

I can't even parse this. What the heck is the difference between information and experience? If experience isn't information, than what is it?

Should we ignore experience because we can't quantify it? Should we ignore it because we don't understand from whence it came? What would that make us?


Just to elaborate on what some have suggested above: it's important to recognize that there's a dialectic between first-hand experience and what you might call "second-hand" information. The information I have shapes how I relate to - analyze and internalize - experiences I have; in effect I can dismiss or downplay my own experience in some cases. (e.g. if I feel chillier this summer than last summer, I hopefully know that doesn't disprove theories of global warming).

And clearly experiences I've had change how I react to information that validates or invalidates those experiences, just as people have suggested above. Nobody I've ever met is a racist via *theory* - they tend to claim their judgements rest on direct (literally irrefutable) experience.

It's interesting too, how one comes by "information" - I've definitely noticed that I'm more emotionally attached to conclusions that I've come to by analyzing information myself (boy, that upsets our neat classification, don't it? - is that experience or info?), than I am to "facts" or hypotheses someone handed to me ready-made.



As many have noted, we have a definition problem here. It seems to me that "information" is a very broad term, encompassing any data that might be considered in making a decision. "Experience" composes some portion of that data set, and is divided up -- there's "my experience" and then there's the experience of everyone else. And then there's other data -- computer generated data, for instance, though even that is probably generated by aggregating experiences in the form of statistics.

If defined this way, it seems to me that it is sometimes best to rely on "information" in the aggregate than on "experience", where you are referring only to your own experience. The experiences of many will likely have more bearing on a good outcome, because they will encompass more data points.

But there seems to be a problem with this general conclusion, depending on the decision to be made. The experiences of many will answer the question "Which car is safest?" better than the experience of one. However, the experiences of many will not be entirely helpful on the question of "Should I enroll in culinary school?" In the latter case, my own experience, and the experiences of people close to and similar to me, will be far more helpful. So the question of "bad information" is really a question of relevant experience -- whose experience is most relevant to the question at hand? Bad information = irrelevant or insufficient experience.


Kevin H

Experience is just another form of information. Presumably you both are talking about personal experiences witnessed with the eyes rather than information you get on paper. If so then my answers would be:

1. Wide range. Academics are certainly trained to try and take information over experience, but even there we all know of some people who refuse to see information right in front of their face because of some past experience. I'd say, however, that the public at large is still driven largely by experience.

2. Again mixed bag. Experience is more prone to outlier effects. For example, witnessing a plane crash might make someone more likely to take a buss than a train, even though unbiased information would tell you that the plane is probably safer. On the flip side, information is naturally reductionist, so you miss out on a lot of subtle context and other non-quantifiable types of information when just looking at stats.

3. Yes, bad information is bad, but so is bad experience. see answer 2.

4. Well, sure, it could happen. Will it happen anytime soon? I'd rather try guess the price of the Dow tomorrow.


Michael Bolton

Suggested reading: The Social Life of Information, by John Seely Brown and Paul Duguid.

B K Ray

Interesting. I think people sometimes weigh one againt the other, but that does not mean that either has an advantage over the other. The dangerous thing about information, is that people tend to collect data to support what they already believe to be true and they tint their experiences in the same manner.

If you are a firm believer that cats are better pets, based on your experience with domesticated animals, you would read Cat Fancy magazine and your trip to the dog show would end in disaster.


Here's the problem with the quote. The people of Brooklyn had predominantly experienced Brooklyn more than New York (Manhattan), and had an identity tied to Brooklyn. The people of Brooklyn would have to look upon themselves with disfavor if they looked upon Brooklyn with disfavor. Here there is a bias that makes the experience unreliable data.

Here's the problem with bad information, the really good bad information never reveals itself to be bad. The idea that Matt Walsh had a videotape of a Rams practice for the Patriots went from something attributed to Walsh to a rumor that had floated within unnamed NFL circles. Or worse, a better alternative for information is never revealed. Every politician is a source of bad information, but to support one requires believing something one is saying on some level. We must knowningly place faith in known sources of bad information to act.



Sounds like one of those great unverifiable claims to me.

It seems that interesting pop-psychology books (I don't read the real stuff to know any better) indicates that we merely claim to make decisions based on information, in reality we still make them on experience for the most part. We just adjust our view of the facts in hindsight to fit what we already decided.

Gordon Haff

My impression was that it was pretty well accepted that people tend to use personal experience (as well as anecdotal experiences from friends)preferentially to data. So, if you have a friend who likes his Ford Ultracar, that will often have a bigger influence than some survey of 1,000 auto buyers. Richard Thaler, among others, has done some work in this vein as I recall.

Jim Farmer

I see it every day. Information overload is a very real problem, especially when folks get conflicting information from "trusted" sources. The easiest thing to do is to retreat to personal experience and "gut" reactions. W sold himself on just it :)


not a chance in the world that your average person in the united states treats information as a substitute for experience. perhaps in europe, but it doesn't happen here. call it the 'blink' phenomenon, or call it the reason a guy like gladwell catches fire at all, but this is not a country that knows how to consume information with anything resembling success.


I think everyone is missing that before the information explosion we have today, often the only information of any kind people had was their own experience. Information gained from experience is just as valid as any other information. After all, who is to say that the source providing the information isn't misrepresenting, misinterpreting or outright lying about it in the first place?


1. Yes - We have access to information about subjects to a degree we could never touch with experience.

2. Yes - It's a net positive although I think we need to manage the propensity to think we know more than we do. (Taleb)

3. Yes - Bad information is a danger. There is "noise" out there that people think is information that leads them to investment choices they shouldn't make. We need a way to test the reliability of information. We do this though a set of expert filters and we weigh the aggregate. Or we ignore almost all of it as taleb does.

4. Experimental wisdom goes hand in hand with information. It's where we get much of it and how the experts sift information for us.