Dynamic socio-political scenario in the post-Nehru era motivated several electoral studies. Kondo (2007)can be credited to have made a review and consolidation of all `electoral studies' of India. One factor which has been found common to all elections after independence is the participation of Indian National Congress (INC). There has been no study which combines the results of the Lok Sabha and State Legislative Assembly elections as determined by various social, political and economic variables. This paper is an effort to study the swings of vote in favour of INC in every quarter under both the type of elections from 1977 to 2007 as determined by various social, political, and economic variables. A multiple regression model has been used for this study. Election results leading to governance of the state or the country by one party or a combination of parties, has very important implications. Hence, the importance of this study.

Indian election scenario has seen three phases. The first phase was post- independence period from 1947-1977 when Indian Congress party was the only prominent national party. There were other parties who were locally dominant. But at national level, Congress was expected to win hands down with no major political, social and economic issues having any impact on election results. The immediate post-emergency era saw the emergence of Jana Sangh as the alternative political force. The BJP of today is a broken fraction of it. The economic and political issues started playing more prominent role in deciding election result. Since the 1990s, after the assassination of Rajiv Gandhi, which was believed to have weakened the Congress party to certain extent, different national political parties started forming alliances before and after elections. Predicting the factors determining the election results have become much more complex. Since independence, only one thing has been common to all Lok Sabha and State Legislative Assembly Elections, INC has been the contesting party in all elections. So, complete analysis of election results for INC has been possible.

Predicting the election results in India is much more complex than other democratic countries due to mere size and diversity. In the USA, for example, study of behaviour of stock market is good enough predicting factor for determining almost accurately the election result. In India, there is no single factor which is actually a good indicator of the election result. The 2004 general election was a real eye-opener. The BJP led NDA government was riding on its slogan of `India shinning'. All predictive models and all opinion polls were indicating a landslide victory for NDA. Gross Domestic Product (GDP) growth rate was high. There was no major communal/caste issues on air. But the political analysts fail to see that more than 80% of our voters live in rural areas. They are not much aware of the growth of GDP or development in the IT sector. To them what matters is the day-to-day survival. Many states were starving due to droughts. Many peasants in the remote villages of Andhra Pradesh and Karnataka had committed suicides. The wealth generated by the boom of the IT sector on the two states had not percolated to the remote villages. Moreover, the Supreme Court, for the first time, had made it compulsory for all the candidates to reveal their personal, educational, financial, legal and political details. Nagadevara (2005) had used classification model and neural network model to study the impact of candidates' personal background in determining the election result of Karnataka State Legislature. The prediction model was found to be accurate to a level of 98%.

Governance And Public Policy Journal, Indian National Congress, INC, Election Results, Multiple Regression Model, Stock Markets, Gross Domestic Product, GDP, Consumer Price Index, CPI, Economic Policies, Political Parties, Post-Emergency Era, IT Sectors.