Some voters may decide between parties and issues on the advantages of their substantive contents by operating under a rationale that is basically individualistic. However, the collective macroscopic voting patterns are the main determinant of electoral outcomes and this points to a different conception of the democratic process. On one hand, the measure of entropy imposes the holistic view of the set of election alternatives. The individual voter has only a single choice in the set of event even though they allocate their votes on the basis of how other voters are likely to cast their votes.
It may therefore be said that the voter is voting for the set of events and not for any specific event in the set. For the necessary distribution of votes to be achieved among alternatives, voters make choices that they would not make in isolation or non-collective choice situation. In other words, voters are not making rational choices from the view of an individualistic logic or individual preference categorization of the alternatives.
The predictability of voting behavior is related with many social and economic factors and it can be expected that any particular election or social system will have an impact on the variation in voting behavior. Since these factors vary from one society to another, they may not be employed in giving general explanations of voting behavior. Past studies of elections have either rejected or modified the traditional ideas about elections and social choice (Dalton & Klingemann, 2007: 246).
However, these studies have negatively done this through contradiction and failure and have not brought forth alternative general, predictive, nomic or even quantitative theory of electoral behavior. The basic measure on an interval scale has been generated by the utility approach so far. One may conclude that this assumption is of minimum relevance to actual behavior in the context of social choice since all the studies integrate the principle that personal preferences dictate the way that alternative social choices are to be ordered.
Once this idea is rejected, one is naturally predisposed to find a measure that does not depend on the specific content of events or their relative preference. Numerous research findings support this view (Coleman. 78). According to rationalistic voting theorists, individuals do not posses sufficient education, knowledge and other skills which may allow them to make rational choices. It could be added that if voters had sufficient knowledge to make decisions, then they should behave rationally.
This perspective may also be rejected when interpreted from entropy. One may arrive at a conclusion that voters actually know what they are doing when elections are taken as entropy measures and result of entropy hypothesis test is accurately viewed (Clarke, & Stewart, 1994: 1104). With this regard, it may be claimed that voters have made rational decisions that are informed by the degree of uncertainty in their society. Rationality is not considered here as individualistic rationality but rather as a collective rationality.
However, the decision of whether individual voters should abandon their collective rationality and vote according to personal logic is also an issue worth examining. This would need many people to act in ways that go against their natural social tendencies which is an unlikely prospect. This does not however mean that the outcome of elections is entirely predetermined by the level of entropy when at an equilibrium. The event probabilities may vary if there is a possibility of two choices. Again, a redistribution of votes may occur if there is a change in the number of event in the set of events.