“In this book, we begin with a simple question: Do groups usually correct individual mistakes? Our simple answer is that they do not. Far too often, groups actually amplify those mistakes.”

If you want to find out how an appellate judge will vote about an ideologically connected case, many of us would check if they were appointed by a republican or democrat. That’s a pretty good predictor. But in many areas, there is an even better predictor: who appointed the other two judges on the panel?

Being in a group affects what decisions we take. That much is clear. But should we take a decision alone or with a group? Both methods seem to work in some situations, and not work in others. In Wiser, Sunstein and Hastie set out to explore the difference.

In general, using a group does two things. It increases the cost of making the decision, because all group members have to be consulted and negotiated with: and it reduces the cost of errors, because it means the group can aggregate wisdom and ideas to minimize expensive mistakes. Or at least, that’s the theory.

In practice, groups can help or hinder. At their worst, they can amplify the errors of some of their members, get trapped in groupthink, become more polarized and extreme without become more correct, and focus only on shared information rather than the key information only some members possess. Perhaps the biggest concern is happy talk: group members go along with the consensus or say nothing is wrong, instead of providing new, sometimes critical, information. Good group design—red teams, giving equal voice to members, and other simple methods—can minimize these costs.

At their best, groups can guess how many jelly beans are in a jar. Groups are good when you need to forecast an unknown result, whether a presidential election or the number of jelly beans: they can aggregate opinions and if everyone is slightly off, the final result can often be very close because the average minimizes the individual errors.

Though Wiser discusses these themes at length, it is short on clear lessons, and indeed the book itself can feel repetitive or meandering. It is an interesting idea, and an important subject, but not as well or as clearly explored as I would have liked, making it hard to distill or learn from.