As evidenced by our acceptability study, reported in
Chapter 
, acceptance of DSV depends not only on
proof (or convincing arguments) that DSV is theoretically better than
traditional systems; there are psychological, political, and
sociological aspects that must be considered along with objective
scrutiny of DSV. We now examine issues relevant to the acceptance of
DSV for polling and voting. Our approach is first to examine those
conditions that would imply eventual acceptance of DSV, and then to
examine the likelihood of those conditions occurring. The discussion
here is broadly based on the theoretical and normative properties
associated with DSV as well as historical precedent and context that
have heralded the acceptance of other voting systems.
Our analysis in Section suggests that people are most
likely to adopt new voting systems at times of great change or
following elections with perverse or undesirable outcomes. In
elections for which plurality rule is used, such perversity often
occurs when more than two alternatives are under consideration. DSV
might be considered as a remedy in such situations. Although
governmental elections may raise awareness about the manipulability of
traditional voting systems (for example, the presence of third party
Presidential candidates tends to promote discussion about whether it
is wise to vote for a candidate with little chance of getting
elected), DSV is unlikely to be introduced for these elections without
first becoming well established as a system for non-governmental
elections. Organizations with educated members who must frequently
choose among many alternatives are likely to be among the first to
adopt DSV.
Our acceptability study suggested that, at least for a well-educated population, voters are open to the idea of a new voting system and the concept of voting strategies. However, it also demonstrated that education about DSV and voting in general will be essential before DSV will gain acceptance. The aspect of DSV that raised the most objections among our survey respondents was instability. Except in situations where a vote is for advisory purposes only or where voters are willing to accept an outcome based on some randomness, the instability of ballot-by-ballot DSV is likely to prevent its acceptance. If stability can be improved or presented in a way that makes it more palatable, or if batch DSV can be improved, the acceptability of DSV is likely to increase substantially. We believe this factor is significantly more important for acceptance than DSV's violations of some of the normative properties discussed in the previous chapter, as the degree to which traditional voting systems possess these properties does not seem to be indicative of those systems' popularity.