Decision Rules

Group decisions are driven by rules—constitutions, bylaws, contracts. Often the set of choices voted on by the group has been winnowed down by a committee or a backroom process that can strongly control the outcome by determining what choices are offered (and how they are described). This prescreening is often filled with obscure rules and processes. Organizations that come to crucial decision points (sometimes vital to the organization’s future) may find themselves suddenly looking at their bylaws (or whatever controls these processes) to find out how things should be done, but when those rules are poorly constructed (or give immense power to a few select people within the group), bad decisions can emerge that please very few people. The time to review organizational bylaws and rules is before crucial votes appear, not in the midst of major decisions themselves.

Poll Vaulting

This chapter provides a set of real-world examples of how the process used for decision-making dramatically affected the outcome and shows how different voting rules would have led to different choices. Examples include the 2000 U.S. presidential election (Bush vs. Gore, with Nader intervening) and the choice of finalist candidates in the 2016 U.S. presidential election (Clinton vs. Trump). It also includes a famous vote by 11 wine judges in France in 1976 (sometimes called “Judgment of Paris”), where we have the actual preferences of the judges. American wines won the blind taste-testing, shocking the French and eventually leading to the “democratization of the wine world . . . a watershed in the history of wine.” This chapter shows that even votes by small numbers of people can have significant effects and that the choice of voting method in part determined which wine “won” the contest.

Systems of Complexity

Every day people make decisions, from trivial to critical, just as organizations. Groups making decisions commonly have some rules guiding their approach—bylaws, constitutions, or just custom. Each approach can be vague and subject to different interpretation and also manipulation. This chapter starts off our discussion about how to make better choices, individually but primarily in groups where the preferences of multiple individuals must somehow be combined to create a collective choice. Decision making on consequential matters often come down to making tradeoffs. Becoming better off on one dimension might mean becoming worse off on another. These decisions are innately matters of sacrifices and compromises. Both systems analysis and social choice analysis focus on how to make these types of tradeoffs. These issues form our goals for this book—how to help organizations make better decisions, by focusing on all three of the key issues: who gets to decide, how do they decide, and what should they decide.

Choosing How to Choose

“Majority rule” seems like the fairest and best voting process imaginable, yet it simply doesn’t work in many cases. Dozens of different voting rules have been proposed to assemble the preferences of individuals into a collective group decision. Some use rank-order ballots. Some use rating scales (like “stars” for Uber drivers or amazon.com purchases). Some use letter grades like those assigned by teachers to students. This chapter reviews over a dozen different voting methods, showing how and why they can lead to different choices even with the same set of voter preferences using a simple example where 19 people (e.g., at a conference together somewhere) vote to decide where to have dinner—an American, a Brazilian, a Chinese, or an Indian restaurant. The powerful conclusion emerges that it is the voting method, not the voters’ preferences, that determines the choice.

Value and Values

Many decisions are done intuitively. Sometimes, this works well, and sometimes they lead us astray. Tools of systems engineering recognize human biases and ask about what we do best—specify what is most important to us under the circumstances. This chapter presents a brief introduction into this world (multi-criteria decision analysis) using an example comparing three fictitious wines to show how different preferences lead to different rankings of wine quality, even when using the same “objective” data. This is as it should be—tastes differ, and good decision support systems take this into account. At this point, we focus on decisions made by one person. Later chapters focus on combining a diversity of individual preferences into a group choice.

Human Factors of Democracy

This chapter takes a new look at different voting methods, asking not how they work but rather how well they let voters express themselves. The most widely used method in the world (“choose one candidate”) is by far the worst at allowing voters to express themselves. This type of ballot has the vocabulary span of a six-month-old infant. Widely used rank-order ballots are modestly better (about the vocabulary of a two-year-old child) but are still very weak as communication devices. New forms of voting, including range voting and majority judgment (where voters grade the choices), offer vastly more ways for voters to express their true sentiments about the choices offered to them. This chapter also assesses how well voters are likely to understand exactly how voting systems work, possibly affecting their trust of the process.

Insincerely Yours

Every known voting method can be manipulated, hence the tactic of “strategic voting.” But some voting methods are harder to manipulate than others and therefore encourage “truthful revelation” of voters’ preferences. These interact with the problems that various voting systems have (following from Chapter 4) when voters are asked to choose between more than two options. Many organizations try to deal with the “problem” of many choices by breaking every decision down into a set of “yes/no” votes. The most famous of these “parliamentary process” procedures is Robert’s Rules of Order, widely used throughout the world. Controlling the sequence of “yes/no” votes can often manipulate the outcome, which is known in political science as “agenda control.” Other ways to manipulate outcomes includes how the choices are described (“framing” effects) and the biases (either explicit or implicit) of “decision facilitators.” This chapter discusses these issues and suggests ways to avoid their being used adversely in group decision-making.

Epilogue

Two major branches of social and engineering sciences—social choice theory and systems analysis—seldom come together. Social choice theory guides the methods we use to choose among a given slate of alternatives. Systems analysis helps us think of new choices. Our efforts in this book are to show how the two areas of thought can be combined fruitfully to enhance organizational decision-making. Systems perspectives encourage thinking about how things interconnect. Deliberation processes that allow nonlinear thinking encourage systemwide thinking. Multi-criteria decision methods tell us how to evaluate choices with multiple dimensions of value. New methods of social choice give people methods to vote on all the choices together, holistically. Combining all of these will lead to a system that produces better design, decisions, and democracy.