Elections, Volatility

A recurring exercise in the field of electoral behavior is assessing to what extent parties' share of the vote and voter choice remain stable over time. Voters may remain loyal to the same party they voted for in the previous election or change preference by voting for another party, abstaining, or spoiling their ballots. A significant number of studies have been devoted to understanding the dynamic of such changes in contemporary democracies. The change of voters from one party to another in the election is called election volatility.

Possible changes in voters' preferences over time may be measured on the individual level or on the aggregate level of the electorate. A voter may change preferences during an election campaign, or over a longer period involving at least two elections. The traditional way of identifying the level of permanence/change in electoral preference on the individual level is quite simply asking voters. For this reason, election surveys are the fundamental source for picking up permanence/change in preference on a microlevel.

For electoral studies on the aggregate level, the source of data is the final result of the race (parties' vote, turnout, and spoiled votes) for a certain electoral unit (national or subnational) in at least two elections. Easy access to election results throughout the world may be the main reason why studies about electoral changes on the macrolevel have been conducted with such frequency over the past few years.

Many indices have been proposed to measure permanence/change in electoral preferences. The most fortunate, and one widely used at present in studies on comparative party systems, is the volatility index, created by Mogens Pedersen (1979, 1980). The term volatility first appeared in chemistry and refers to the tendency of a substance to vaporize. It has also been defined as a measure of how readily a substance vaporizes. The notion of a change in physical state captures the fundamental idea of the index, which is to quantify the intensity of change in party preferences when two elections are compared.

The Index

The volatility index (or total volatility, TV) measures the level of aggregate electoral change over two consecutive elections. Although the index is traditionally calculated for the total number of votes received by parties, it can also be calculated including total turnout and spoiled votes. Its formula is

where V represents the percentage difference of votes (or seats) for each party in two consecutive elections.

The volatility index is calculated as follows: The percentage of votes (or seats) that a party received in an election is subtracted from the percentage of votes obtained by this same party in the preceding election; the difference indicates the change, and the (–) or (+) signs reveal the decline or growth of a party, respectively. The next step is to add the result of this operation (not considering the sign) and divide by two. The value of the index expresses the total number of votes lost by the parties whose votes fell from one election to the next or the total number of votes gained by the parties whose votes grew in the same period. For example, a volatility index of 10.0 indicates that the combined share of the vote of all the parties whose vote fell was 10 percentage points and, consequently, that the parties whose vote grew, taken together, increased their share of the vote by the same figure.

...