A Pew Research Center survey conducted between December 7, 2013 and January 12, 2014 based on face-to-face interviews with 2,464 randomly selected adults and a margin of error of 3.8% says:
More than six-in-ten Indians (63%) prefer the BJP to lead the next Indian national government. Just two-in-ten (19%) pick the Indian National Congress. Other parties have the support of 12% of the public. BJP backing is consistent across age groups. And support is almost equal between rural (64%) and urban (60%) Indians.
Surprisingly, however, the survey shows the following breakdown of the BJP's support-base, which shows it doing even better in the South (where it has practically no or very little presence in Tamil Nadu and AP, and Kerala is not sampled) than in the West:
Perhaps the answer to this lies in the methodology (Kerala was excluded in the South, see below) and voters were only asked which party should lead the next government:
The Pew Research survey does not directly ask about vote choice or the likelihood that a respondent will vote, and it cannot gauge the level of support for either the BJP or Congress in particular constituencies. But it reveals a widespread desire among Indians for a change in leadership.
Likewise, when it came to the comparison of the prospects between Rahul Gandhi and Narendra Modi, the poll shows the following:
Interestingly, while Arvind Kerjirwal was not included in the list of leaders whose popularity was sought to be gauged, the leader who came #2 in popularity to Mr Modi was Anna Hazare and even Mr Manmohan Singh did better than Mr Rahul Gandhi:
Results for the survey in India are based on 2,464 face-to-face interviews with adults 18 and older, between December 7, 2013 and January 12, 2014. Interviews were conducted in Hindi, Tamil, Bengali, Telgu, Odia, Marathi, Kannad, and Gujarati. The survey covers 15 of the 17 most populous states (Kerala and Assam were excluded) and the Union Territory of Delhi, which together are home to about 91% of the adult Indian population. The survey is based on an area-probability design, which entailed proportional allocation of 1,876 interviews by region and urbanity, plus an urban over-sample of 588 interviews. The primary sampling units were urban settlements and rural districts. The full sample was weighted to reflect the national urban-rural distribution in India.
Read the full report at Pew Research Global Attitudes Project