Results of Bruce Schneier's experiment in trust
schneier.com
schneier.com
Phrased that way, the statement is absolutely incorrect. What is very likely is that a great majority won't uphold their end of the bargain. What is also very likely is that few if any of those made a decision not to uphold their end of the bargain.
Anybody who had that intention from the outset would have simply torrented the book. Anybody who didn't have that intention from the outset is unlikely to switch intentions. This is known as "consistency bias".
But what does happen is that people make commitments and then forget about them if the stakes are low. That is almost certainly what has happened in virtually every case of noncompliance in Schneier's experiment. I have a lot of respect for Schneier and his security and cryptography work. But his science fu is revealed in this article to be rather weak, as is his understanding of human psychology.
I disagree. Although I agree that it is unlikely that it was their intent from the outset, there is a threshhold at which point someone has practically decided not to do something. Once they have procrastinated on doing something for X months and know that it is only #73 on their internal priority list, they have essentially decided not to do it. It could happen, but the odds are, what, maybe 10%? If you assume that a commitment to do something is giving a 100% guarantee and a commitment to not do something is a 0% guarantee, a 10% chance is, pragmatically, a "no".
Also, part of the implied deal was that you write the review promptly. From the original blog post where Schneier made the book offer:
> Liars and Outliers has been out since late February, and while it's selling great, I'd like it to sell better. So I have a special offer for my regular readers.
I don't think he was hoping they would send in their reviews a year or two later. Those who do would have a hard time arguing that they met the spirit of the deal.
The initial blog post was almost exactly 5 months ago and the offer was closed in two days, which means that a lot of readers probably got their books at least 4 months ago. (In my experience he does ship promptly, as I have gotten a book from him before.) If he was hoping that they would write their review in a reasonable amount of time, say 8 months, and only 9% wrote a review in the first half, I highly, highly doubt that even 41% will write a review in the second half. And the majority of those who will not probably subconsciously know it by now.
Edit: To be specific, I don't think he expected 90% turnout by this date, he's just disappointed that the turnout is so low by now. It's not even 1 in 10. 4 months is a decent amount of time to review a book you agreed to review.
I think the point is that a decision is an affirmative act. Nobody decides to forget to do something. It's like the difference between incompetence and malice.
But didn't they choose not to ensure they remembered?
That sentence sounds ridiculous. But if I care about something I'll ensure that I get reminded somehow so I can't forget. The experimental subjects that didn't provide a review either explicitly chose not to, or acted in a way that made it possible for them to forget.
The same sort of argument is made by people who protest a vote by not voting. Not voting doesn't convey any information as to the reason for not voting; it's lumped in with "didn't vote because of apathy" or "didn't vote because of illness". There's no way to infer intent from a non-action.
So how do we make conditional commitments? Sometimes it annoys people when I won't answer yes or no to a commitment but instead say "I'll do my best." When pressed I can articulate the various uncertainties between now and the execution of the commitment that could derail it, but they are just that uncertain. Sometimes people say "Just say yes, and if something comes up let me know." I've done this in the past and had people accuse me of 'making excuses' which of course I feel is unfair :-).
There is certainly an element of 'value' here in that the consequence of not doing something has a cost and if the cost is small sometimes less effort is made in the follow through. If money is involved that is generally a more explicit definition of cost. But if someone says, "Hey my son Jimmy is on the special team at this weekend's High School football game, can you come by and cheer him on?" what is the cost of being unable to do that? Sometimes small (there are lots of people at the game) and sometimes large (this person really wanted to see you validate their support of their child's athletic activities) and sometimes in the middle.
Schneir positioned the transaction into the 'trust' category, you say you will do something and I trust that you will do it. The consequence of you not doing it is a loss of trust on my part. But the trust currency isn't previously established. Clearly my kids have a vested interest in keeping my trust level high because I can restrict access to the vehicles, but Bruce doesn't know me, and if he trusts me more or trusts me less do I care? Do I care $29 worth? Or more importantly do I care 'marginal value of Bruce's book minus $11' worth?
As you can read its something I wonder about but have few really solid lines of reasoning about. A lot of questions, a lot of speculation about the intrinsic forces in the 'trust market' and no clear experimental path.
Without debating the orthogonality of that assumption, (because it really won't change the overall point), let me begin by saying, most if not all humans I've encountered, have varying levels of principles and opportunism, but we are not made equal (CF fallacy of gray).
>I won't answer yes or no to a commitment but instead say "I'll do my best."
I am one of those, that have traditionally found that statement slightly annoying, especially in a business context. It just seems slippery. What does "doing your 'best' mean? Are you truly going to do your very best and drop everything to fulfill the request? I think not. Hence the precision one is aiming for, does not seem like an improvement on the original brevity of the yes/no.
Furthermore it could give one an opportunity to flake and blame the environment, versus sticking to one's word.
I completely understand this response, I actually really do mean I will do everything I possibly can to meet this obligation, but also recognize my commitment is not equivalent to other commitments by people who might use the same phrase.
I find myself stuck in this dilemma:
1) I can say "no" which cuts off the conversation and makes me appear unsupportive.
2) I can say "No, I don't think I will be able to." Which at least invites the requestor to inquire further.
3) I can say "I'll do my best" which accurately conveys my personal intent but sounds weak to the recipient.
4) I can say "Sure, I can probably do that" where I signal that my intent is yes but I can see issues which may interfere with my execution.
5) I can say "Sure" and 'bet on the come' as they say in Vegas which means that if things outside of my control happen in the best possible way I'll be able to fulfill the obligation, and if they don't I'll fail at filling the obligation and let you down.
I have observed that option 5 is the normal usage. It results in experiences like the one Bruce had where there are lots of commitments and not necessarily a lot of follow through. And at low "value" in terms of the commitment/cost the discussion of how much trust is impacted is interesting.
I've also observed that people in management that always say "yes" even if they only deliver a smaller percentage of the time on their commitments are more successful (in terms of raises, promotions etc) than people who actually only commit to things they actually think they can do[1]. So as a success strategy always saying yes, regardless of your understanding seems like a winning strategy. I personally can't do that though, I'm sure it has hindered my career :-).
[1] There is also a very real thing that you can often do more than you think you can, so sometimes over-comitting gets you to a higher level of efficiency. It still annoys me when someone signs up to do something that I know for a fact they won't be able to deliver, and they get full credit for the signup anyway.
But a contrary view is that the holidays are when most people are on vacation, and hence more likely to read a book and have time to write a review. Assuming they had time to read the book, it's not like the review requires a considerable time investment (comparatively).
There's lot of research into getting people to make and fulfill commitments.
If for example he had made everyone write out "I promise to write a review of this book within two months of receiving it and post said review online", compliance would have been much higher. If they had been required to post that promise publicly to their friends when they made it, compliance would have been higher still.
>My only request is that, after you read the book, you post a review about it somewhere. On your blog, on Amazon, on -- I suppose -- Twitter. Just let people know about it.
The final sentence reduces it to a "make sure you tell your friends about us", which one usually responds "Ok. I will" but doesn't really plan to do so.
I am glad to see though that he did publish lists of the reviews he did find so that anyone whose review was overlooked has a chance to respond.
Unfortunately, his post there ruins the experiment.
I did not participate in the experiment but if I had and was about to go live with my review and then suddenly was publicly accused of being a thief, I would not go forward with an ordinary review publication since it would seem at that point to be trying to wiggle out of something. Instead I would write a blistering criticism of his methodology.
The claims and article are so shoddy in fact that it damages my opinion of his work overall.
I sometimes write book reviews on my blog. I don't accept free or discounted books, it's just books I've read. I have quite a backlog. Right now I have a stack of 30 books that I will read. Some of these I won't get to for a couple years. Not every one of them will merit a review. It's not reasonable for an author to expect amateur reviewers will get to their book within a few weeks of receiving it, especially when there is no stated deadline.
Additional thoughts:
* Reading further through the comments on his blog, in the first few comments I found 15 people who mention they are in the process of reading it, or who have not gotten to it yet but plan to. There were more; I just stopped counting at 15. There are no comments (among those first ones at least) from people gloating that they got away with an unwarranted discount. Several fulfill their obligation inline by posting their review thoughts right there in his blog.
* It should also be mentioned also that these days even hardcover books in modest quantity are very inexpensive to print when there's no color plates, so the $11 paid almost certainly ensured that money was not lost on the deal as far as printing and shipping costs goes.
* Deep down in the comments Scheiner responds to the comments and it seems he realizes he screwed up with his post and drastically underestimated the number of people who take a while to get around to and complete reading a book.
I think that's a little more than he concludes. He is always thoughtful, which is a good reason to read him, and he takes pains to say at the end he wasn't complaining only noting. (I don't know if you posted before he added that.)
If I can give my $0.02, coming from my perspective as having worked on a free archive for a long time. My personal experience with giving to people "out there" is that you can expect little return, beyond personal gratification. However, on reflecting on my experience, I've decided that I can't say that the people are freeloaders; probably most of them coach little league, or help out in meals on wheels, or shovel the walk of the old lady down the street. But everybody is very busy and while they sincerely intend to be part of the pay-it-forward group, on any one thing the return rate is just low. Myself, I do the same stuff.
I think this is Bruce Schneier's way off the mark, even inappropriately so. His book culture is not everyone else's and he shouldn't by default expect that from them.
Paying $11 for a book and then having to do a review is not an amazing deal for the buyer. It's just a deal. If no time line was specified in the deal people have every right to read the book when it suits them. They didn't save $5-$15 to get a chore.
For me that would have been a within a year estimate.