No news is good news? Think again

A proverb that never seems to die is the oft-heard, “No news is good news.” However, a new study published in the American Economic Journal: Microeconomics suggests that this not only goes against logic but also reduces returns in an information sharing game.

A marketplace for information

According to the principles of information economics, most firms have an incentive to release information on their products if the cost of doing so is low. The belief is that customers will treat firms that don’t release information as being the same, so the one that does provide information — say, about the quality of the manufacturing — looks better in comparison for doing so. Over time, this should lead to more information being disclosed voluntarily as people try to cash in on the effect. An “unraveling” effect also occurs, with customers assuming the worst about those keeping secrets.

However, like a lot of ideas in economics, this one rests on the assumption that consumers will behave rationally or according to theory. Sometimes, people just don’t catch on to the fact that some businesses might have an incentive only to reveal good news or to brush unflattering details under the rug.

To help shed light on how people actually behave and why they do so, the researchers set up an experiment based on an information sharing game where participants could win cash prizes for using information economics to their advantage.

An experiment in information economics

The main set of experimental sessions involved two randomly matched players, one as an information sender and one as an information receiver. The sender would be given a secret number between one and five by a computer. They then had a choice to reveal this number to the receiver or not. Lying was not allowed.

The receiver, who either saw the number or a blank space on their screen, then reported what they thought the secret number was. While the senders always saw and submitted whole numbers, receivers could guess any half unit between 1 and 5.

Sender players earned rewards as receiver guesses went higher, no matter what the secret number actually was. Receivers earned more for accuracy, with perfect determination of the secret number getting the most money.

The experimental set-up matches the theory; namely, the best “moves” are for the sender to always show the number unless it is one, and the receiver should always guess the number is one if they don’t see a number.

Theory meets reality

But, things get muddy when theory meets reality.

After 45 rounds of play, senders tended to be between 3 and 7 percent off from the highest possible payouts, while receivers were 9 to 13 percent off. Numbers that you would expect to have been shown by the senders remained secret. Receivers made strange guesses, sometimes not trusting the values they saw (despite being told lying was against the rules) or guessing higher than they should have when shown a blank screen.

The authors mention that the receivers often seemed “insufficiently skeptical of nondisclosure” and failed to realize that the sender was probably hiding something from them. This effect could be somewhat mitigated by providing them feedback, though it had to be offered repeatedly for their improvement to be sustained.

They also opined that some receivers might have been confused by the rules of the game and played poorly as a result.

For the first time in a study like this, the researchers also asked the senders what they thought about the receivers. Their choices to reveal the number or not were often driven by a belief that the receivers would respond to not getting any information by guessing just below the middle of the number range rather than the lowest value as they should have.

As it turns out, they were right, with the average blind guess being above two. While the senders were not following the theory by acting this way, neither were their opponents — and thus, they were playing the game optimally.

It seems you really should play your opponent and not your hand.

Market forces are insufficient to close the information gap

The authors summarize the possible real world application of these findings in their report:

“These findings suggest that unless buyers receive fast and precise feedback about mistakes after each transaction, market forces can be insufficient to close the information gap between sellers and buyers. For the products that naturally offer such feedback — say cereals that taste crunchy and t-shirts that hold color fast — voluntary disclosure may converge to the unraveling predictions after a buyer purchases the product many times. However, for product attributes with less immediate feedback — such as the fat content of salad dressing and the cleanliness of a restaurant kitchen — voluntary disclosure may not converge to the unraveling results. In these situations, mandatory disclosure may be necessary if the policy goal is complete disclosure.”

Oh, before you go, I promise that I didn’t leave out any information from the study. Nothing important anyway.