The trend that becomes visible from these surveys is favourable to Fidesz and LMP: neither MSZP, nor Jobbik could substantially extend their supporter base, moreover, the support for both parties have narrowed while elections are approaching. The reason for that could be the reappearance of Ferenc Gyurcsány, the continuation of corruption cases connected to the government, and the scandals which question the seriousness of Jobbik.

The potential support for MSZP fell from 22 per cent measured in January to 15 per cent, for Jobbik from 16 per cent to 11 per cent (here, the shrinkage is scenic compared to last week). Thus, probably neither of the parties is able to reach a 20 per cent result and win one million voters. Fidesz was able to retain the voters who were winking towards Jobbik, and those who search for an authentic party on the Left, placed confidence in LMP. The admitted support for the green party with a presently doubtful program grew for the first time over 5 per cent (their potential voter base moved between 7 and 9 per cent for a couple of weeks now). MDF could not profit from the Left wing voters searching for a party: their potential support is mildly below the parliamentary curb (4 per cent).

Estimation of parliamentary seats - In the last research before the elections, we have again attempted to model the constitution of the next parliament. Fidesz-KDNP still has a big chance to get a two-thirds majority (284 mandates), and MSZP seems to be entitled for becoming the "second power" in contrast to Jobbik (50 versus 37 mandates). However, we have to point out the boundaries of these models. In the Hungarian election system, even fragments of per cent can decide mandates. The regional division of votes and participation, and the amount of difference between parties can influence the final distribution of representative seats.

Sample and Method - The interviews were conducted between 31st of March-2nd of April 2010 by telephone on a nationwide representative sample. N=1000. The margin of error is maximum plus or minus 3.2 percent, with different margins or error for subgroups.