In most developed democracies, an important component of a pre-election analysis is to look at changes in victory and defeat margins for major political parties. The underlying assumption in this type of analysis is that there is stability within political parties, a certain level of consistency in their respective constituencies and a high degree of certainty in the larger political system. It also assumes that many voters maintain a fairly stable set of electoral preferences across different election cycles while many others remain undecided till the very end of an election campaign.

These undecided voters are usually swung one way or the other by the economic or social policies (taxation, investment in certain social services, job creation, etc) that different contenders for power promise to follow. In constituencies where victory (or defeat) margins between two contenders are narrow, this swing proves decisive in determining the outcome of elections.

Democracies in developing countries are generally much less stable, with many unpredictable factors playing crucial roles in determining the outcome of their elections. This is mainly due to instability in party structures, volatility in voter turnouts and a general uncertainty in economic, political and social spheres. Consider how the issues that determined voters’ behaviour before and during the 2008 elections in Pakistan (Benazir Bhutto’s assassination, suicide bombings, intensification of public anger against Pervez Musharraf, etc) changed entirely before the 2013 elections. The issues that had a major impact on the last general poll included the almost sudden rise of Imran Khan’s Pakistan Tehreek-e-Insaf (PTI), Nawaz Sharif’s return to electoral politics after 16 years, serious security threats and a large-scale resentment, particularly in Punjab, against the government of Pakistan Peoples Party (PPP) led by Asif Ali Zaradri.

The sheer number of factors that may impact voters’ preferences and their willingness to come out to cast their votes for the 2018 election renders the predictive power of a swing analysis uncertain. It is, therefore, important to look into some of these factors in order to carry out a meaningful analysis of victory or defeat margins and swing votes as a means to forecast the likely outcome of the upcoming elections.

This excerpt is part of the Herald's pre-election July 2018 special issue. To read more, subscribe to the Herald in print.

The write is a senior research associate at the Collective for Social Science Research. He has conducted extensive research in the fields of social protection, macro-economic policy, poverty and corruption.