Most polls suggest that the centre-right coalition will win today's election in Sweden. If this happens, it will confirm that successful policy doesn't help win elections: the SAP managed to pay down the colossal deficit left by the right in the early 1990s, reform the welfare state and push unemployment down, all whilst remaining one of the most egalitarian countries in the world (although there was a significant rise in inequality, from a very low level). Then the right comes along and pushes the same line of policy a little further, and manages to convince everyone they have dragged Sweden out of the abyss.
This suggests a number of pessimistic conclusions for those on the left. First, successful policy may be either disregarded by an ungrateful electorate, or simply misunderstood. After all, if political scientists and economists have such trouble pinning down the exact effects of policy changes, why should we expect the electorate to assign blame and credit correctly at election time?
Second, saving the welfare state is no guarantee of electoral success. First, because the right is clever at talking about reform whilst steering clear of unpopular measures that median voters would object to (Sarkozy seems to be an exception here, though he seems to be playing the distraction card to get his pension reform through). Also because voters are not always very aware of the indirect costs of welfare reforms that may not affect the middle classes directly, but indirectly get them in the end (see Wilkinson and Pickett's The Spirit Level for further details).
The left is trapped in a dilemma between stubbornly defending social rights, even when this is economically inefficient, and embracing the market at the expense of higher inequality. Is this really the only choice we have?