(YES) - Report 4: 6. Party Identity and the Propensity to Vote

Updated: 24 January 2011

6. Party Identity and the Propensity to Vote

In previous reports from the YES project, we have identified a number of different factors related to Australian youth's propensity to vote. Among these has been gender, confidence regarding voting, and so on. It is possible that any one of the factors that we have discussed is only partly linked with voting; that the correlation between factors tends to obscure the full picture of what explains the attitudes and behaviours of youth towards voting.

In this context, we ask the following question: is a person who has a party leaning (or party identity) more likely to say they will vote? But from the other YES reports we know that other factors are also related to the tendency to vote. Therefore, what we really want to know is whether having a party leaning, or party identity, related to the intention to vote, even when the other factors are taken into account.

In order to examine this question, the various factors which already have been found to explain the propensity to vote from the YES survey in a multivariate model, will be used to examine the relative importance of political party identity (Saha 2005).

Using the multivariate regression model from the previous analysis of the determinants of voting, the additional variable, party identity was added. This variable consisted of a simple dichotomy created from the political party variable, and was recoded as naming a party (coded as 2) and not naming a party (coded as 1). The first category of students was considered as having a party identity, while the second was considered as not having a party identity.

The voting variable that is used in the analysis, is commitment to voting, which is a combination of the two variables, 1) whether a person will vote at 18, and 2) whether a person would vote at 18, even if not compulsory. There are five categories in this variable: 1) definitely will vote, 2) probably will vote, 3) Maybe will vote, 4) probably will not vote, and 5) definitely will not vote. Each category is given a value ranging from 6 to 2, with 6 being allocated to "definitely will vote". [2]

The same selection of variables which have been found to be related to "commitment to voting", as reported in Saha (Saha 2005), were used in this analysis. The only difference is that the new variable, "party identity" was added to the analysis. The results of the multivariate regression analysis are presented in Figure 7.

In this figure, the values for each bar represent the value of the standard regression coefficients (or beta weights), which indicate how much of a change in "voting commitment" is related to each of the variables which are listed, controlling for all the other variables. Since the values (beta weights) are measured in standard deviation units, it is possible to compare the relative size of each variable. Those with higher values are more strongly related to "commitment to voting", even when the other variables are taken into account. The arrow points to the variable "party identity".

Figure 7: The Effect of Selected Variables on Commitment to Voting for Australian Senior Secondary Students (Standardised Regression Coefficients or Beta Weights)(Adj R Square = .42)

It is clear that having a party identity (naming a party) is an important variable in understanding youth propensity to vote; it is the fourth most important variable, in this group of variables, in explaining commitment to voting (see the arrow). Only the three variables, namely, 1) regarding voting as important (a civic duty or legal obligation) 2) feeling prepared to vote, and 3) having an interest in politics, have a greater impact on "commitment to voting" than 4) party identity. All the other variables are less important.

The pattern in Figure 7 is consistent with the findings of Howe (2006), in that both a sense of civic duty ("It is important to vote") and party identity are important determinants of intention to vote. Students who could name a party that they identified with, were also more likely to say that they will vote when 18, even when controlling for 12 other variables. This relationship is a strong one, which suggests that among young Australians, party identity alone does differentiate between those more likely to vote from those less likely to vote. In other words, being able to name a party does make a difference in intention to vote.

What remains to be determined is what factors contribute to a party identity among these young Australians. In other words, why did some students name a party that they leaned towards, and other students did not?

Key Points for Section 6

  1. Having a party identity is strongly related to the commitment to vote, even when other variables are controlled. (See Figure 7.)
  2. Among a group of 13 variables, having a party identity ranks fourth in importance in relation to commitment to vote. In other words, it is important. (See Figure 7.)


  1. This variable is a combination of two questions in the questionnaire, namely whether the student intended to vote when 18 years of age, and whether the student intended to vote even if voting were non-compulsory Saha, Lawrence J. 2005. "Are They Prepared to Vote? What Australian Youth Are Saying." in Youth Electoral Study Workshop. Canberra. Thus, by combining these two variables, we have a variable which takes into account the compulsory nature of voting in Australia, and gives us a measure of the student's "commitment" to voting, even if it were not compulsory.