The turnout factor

Print edition : October 23, 1999

DESPITE widespread media reports of a dismal voter turnout owing to 'election fatigue', the overall polling figure for the 1999 Lok Sabha elections was higher than in many previous years. It was higher than not only the 1991 and 1996 elections, but seven previous elections.

Mid-term elections are generally likely to produce lower turnouts than elections held after a government's full term in office. The mid-term polls of 1980 and 1991 both registered noticeable drops in voter participation. Bearing in mind then the fact tha t this was a second consecutive mid-term election, taking place barely one year after the last one, the voter apathy was not nearly as pronouced as might have been expected. Besides the long-drawn-out nature of the campaign and unfavourable weather condi tions in many States, the electorate had many reasons to stay away from the polls. Indeed, seen in this light the turnout of 59.5 per cent shows the commitment of the Indian voter to democracy.

Despite a slight drop in the overall participation compared to the previous election, there were in fact many States in which the turnout increased. As is to be expected, the States that had simultaneous Assembly elections were amongst those that registe red an upturn. With the exception of Andhra Pradesh, where the figure fell a little from 1998, Arunchal Pradesh, Sikkim, Karnataka and Maharashtra all surpassed their previous marks in the 1990s.

There is little evidence of any correlation between the level of the turnout and the success of a party. Despite the media's fascination with the hypothesis that the level of the turnout is directly related to the success or failure of the BJP, the resul ts show that the turnout did not favour either of the two main formations, and that incumbent State governments were just as likely to do well as they were to do badly in states that witnessed a low turnout. In States where the turnout fell, there were s ome major reversals. In Punjab the Akali Dal was routed and in Rajasthan the Congress(I) lost a lot of ground. Conversely, the BJP extended its control over Delhi, Haryana and Madhya Pradesh, all of which recorded a drop in turnout over 1998.

The CSDS post-poll survey data show how turnout figures have varied among respondents from different social backgrounds. The categories are for community, age, sex and locality. The plus or minus figure indicates by how much the group in question is abov e or below the national average. Of the different communities the Scheduled Castes were the most likely to vote, as their turnout was 2.2 percentage points higher than the national average. The community least likely to vote were Muslims, whose turnout w as 2.2 percentage points below the national average. The figures for the upper castes and the Other Backward Classes (OBCs) were roughly 1 percentage points over and under respectively of the national average, but the biggest increases are in the case of the Scheduled Tribes, the figures in respect of whom is a fraction more than the national average, which marks a significant increase in participation by a community that has traditionally been the most reluctant to turn out to vote. Previous CSDS resea rch has shown that there is a participatory upsurge in the 1990s, as citizens from the lower orders enter the political arena. Similarly, the gap between the turnout in the general constituencies and the reserved (S.T.) constituencies has also narrowed. In 1977 the gap was 12.9 per cent, which by 1998 had come down to just 2.3 per cent. The same level was maintained in the 1999 election.

There is a strong age effect at play. The data show that voting seems to be mainly a middle aged activity, with both the young and the old falling well short of the national average.

Men are more likely to vote than women. The biggest differential is found between urban and rural respondents, where the turnout for town and city dwellers is 4.7 percentage points lower than the average, and 7.2 percentage points lower than in the case of rural voters. The same conclusion emerges from an analysis of turnout by the urban or rural nature of constituencies. Since the mid-1980s, the rural constituencies have overtaken the urban ones in the matter of turnout. The gap increased this year fro m about 6 percentage points to 9 percentage points, as the urban centres recorded a turnout of 51.6 per cent compared to 60.6 per cent for the entirely rural constituencies.

Surprisingly, education counts for very little in determining whether people vote or not. Figures for both illiterates and graduates are similar here.

Although there is a certain amount of variation in the social profile of those who vote, there is enough evidence to refute one of the commonly held assumptions about electoral participation. Conventional wisdom from the West states that it is the most d ominant sections of society, such as the educated, the upper caste, the urban, and the men who are the most likely to vote, as it is they who have the highest stakes in the system. This view is supported by data from many other democracies around the wor ld. However, the case in India seems to indicate the reverse, and India is possibly the only country that upsets the pattern. It is now the lower sections of society that have over taken the elite in this respect. The fact that the most disadvantaged mem bers of society have not become disillusioned with the system and still actively participate in political choice-making proves that the democratic revolution set in motion by universal adult franchise still works.