If FPTP can't do what it supporters claim is its primary virtue - deliver a clear choice of governments by awarding a majority to the party with most votes - then it would be difficult to claim it is fit for purpose.
That case is made by ippr today in a paper Worst of All Worlds: Why First Past the Post No Longer Works by Guy Lodge and Glenn Gotttfried, which argues that first-past-the-post is now "broken". The full paper can be read on the ippr website.
I made a similar argument as to why FPTP was "broken" in its own terms in a Fabian Review essay back in the Autumn of 2007:
Few feel any urgency about voting reform. But the hidden truth of our democracy is that the electoral system is broken, and no longer fit for purpose. If elections ever became close contests again, this would reveal that we are playing Russian roulette with British democracy every four or five years.
The analysis underlying that piece is strengthened by the 2010 outcome. While well known to academics, this problem with FPTP has not featured in public debate, perhaps largely because this has mostly been between supporters of PR and defenders of the status quo.
What's gone wrong for first-past-the-post?
The "give a winning party a majority" feature of first-past-the-post is dependent on a happy accident of electoral geography which was in place from 1945-70 (and probably from 1867 to 1906 too), but which has eroded from 1974, and again after 1992.
Not a lot of people know that half of 20th century General Elections before 1945 produced hung Parliaments.
The chances of hung Parliaments in future are now again very strong - as John Curtice's analysis, which the ippr draws on, sets out. This is common sense too: if the Tories couldn't win in the conditions of 2010, what circumstances do they need to win a majority? The common assumption that Cameron would have easily won a second election simply ignores Curtice's evidence as to why the reduced number of marginal seats would have made that considerably harder.
How no man's land grew and grew
The chances of a hung Parliament have again been much greater because of the rise of the number of MPs from neither major party (for which Martin Kettle has coined the term "nottles"). These determine how much "no man's land" there is between Labour and Tory majority territory - with a narrow strip now becoming a gaping chasm:
In the general elections of the 1950s and 1960s, the number of neither Labour nor Tory MPs was 11, 9, 8 and 7 (in the 1950s) and 9, 14 and 12 (from 1964-70). With the fragmentation of the two-party vote, and the increased representation of Liberal and nationalist parties, that rose to 37 and 39 in the 1974 elections, and after falling to 27 in 1979, rose again to 44 or 45 in each of the 1983-92 elections, the jumping again to 76 in 1997 and then 92 in 2005.
The secret of FPTP's survival
What few people have noticed is why first-past-the-post survived - and managed to avoid producing a hung Parliament until 2010.
The answer is simple: we very rarely have close general elections in Britain.
2005 was the first election for over 30 years in which the two major parties were within 5% of each other.
This is lucky - as first-past-the-post probably couldn't survive if we had close elections quite often, which is perhaps the best reason to question whether it can be considered fit for purpose as an electoral system.
The Tories did just squeak a majority when 7.5% ahead in 2992, but their 7% lead in 2010 left them 20 seats short. Labour gained a 65 seat overall majority on 35% of the vote when only 3% ahead in 2005, when an 8% lead would not have been enough for the Tories).
So first-past-the=post has a very high chance of providing not only hung Parliaments, but also bizarre and illegitimate results if the two major parties are within 5% of each other. It picked the "wrong" winning party on two of the six occasions this happened even when the broad electoral geography was in place (in 1951 and Feb 1974). It would tend to do much worse now: in 2010, it could easily have given a third placed Labour party the most seats, and put Nick Clegg a distant third in seats had he bounced ahead of the other two parties on votes.
The government claims its redistricting plans (also shrinking the Commons) will deal with this. They won't. The electoral bias in FPTP has much more to do with the pattern of the distribution of votes than unequal constituency size, which accounts for less than a quarter of these votes to seat differentials. It is part of the system, and can't be ironed out, except through patterns of voting or the party system changing significantly again.
What should majoritarians now advocate?
The academic evidence strongly suggests that those who support the traditional majoritarian arguments for first-past-the-post are no longer served by the status quo - even if politicians still trot out the traditional mantras.
So what might they advocate instead?
Some traditional champions of first-past-the-post, like Jack Straw, think that the Alternative Vote protects all of the features of FPTP they think valuable - the constituency link, and a majoritarian tendency. It is possible to secure quite a broad consensus for AV as the best system to elect a single individual where there are multiple candidates - a position which PR supporters can share with Straw.
AV is a majoritarian system better suited to plural politics, which makes it a good move for some, and a dangerous compromise for others. (Over time, it could prove a "slippery slope" to PR, or it could end the reform argument by delivering a more legitimate majoritarian system. Nobody knows which might happen: these would be arguments for the future). What is unclear is what principled and evidence-based argument supporters of FPTP have to prefer it to AV.
The alternative to AV and FPTP for traditional backers of first-past=the-post would be to promote a majoritarian electoral reform.
It is difficult to see a "winner's bonus" system being considered legitimate - for example, a top-up of 100 seats to the party with most votes, though this is what supporters of FPTP are trying to do, and what the electoral geography now denies them.
The clear and potentially legitimate way to have a majoritarian electoral system would be to directly elect the Prime Minister or Cabinet. That would be a considerable constitutional revolution in Britain - one needs to work out how to elect the Parliament, and whether to have a separation of powers, etc - but it would do something that first-past-the-post can not be expected to do in future: give voters a clear choice between governments.
Can the legitimacy of FPTP be restored?
A referendum victory for the No camp would certainly increase the democratic legitimacy of first-past-the-post, because the electorate would have chosen to stick with it, warts and all.
It is often suggested that this would take electoral reform off the agenda for a generation. Certainly, it will look that way if the "No" camp wins in May. But I suspect things would turn out differently in the end if that's where we do end up.
If it survives this referendum, first-past-the-post is quite probably going to fall over were we to have a close General Election or two - and the public today are going to be a lot less sanguine about picking a "wrong" winner than in 1951. That would give us a significant crisis of political legitimacy.
It really would be better to put it out of its misery first.