THEY may have been overshadowed by the closely fought parliamentary election, but last week's local elections were also groundbreaking. Instead of being elected via a first-past-the-post system, Scotland's councillors were elected using the single transferable vote (STV) method. Rather than placing a single X against one candidate's name, voters were invited to put the candidates in order of preference - 1,2,3, etc. And, instead of electing just one councillor per ward, three or four candidates were chosen using a complex
counting system expected to produce a roughly proportional result.

This switch was expected to transform Scottish local government. Labour benefited from first-past-the-post elections - in 2003 it won 42 per cent of all the seats despite winning just 33 per cent of the vote - while the SNP lost out. If STV had been in use in 2003, Labour would have had around 100 fewer seats, the SNP about 120 more.

But the elections also matter beyond the confines of Scottish local government. The STV system has long been favoured by advocates of proportional representation in Britain, but hitherto its use has been confined to Northern Ireland. However, if its use in last week's Scottish local elections comes to be judged a success, it is more likely to be put forward seriously for use in other elections, including those for the Holyrood and Westminster parliaments.

Scottish local government has certainly been transformed. Labour lost more than 160 councillors. The SNP gained 180. The SNP gained councillors in almost all of Scotland's 32 councils; Labour lost out in 22. Labour not only lost heavily because of the switch to STV, but also its support dropped more generally. On
average across the eight councils whose overall results are detailed in our graphic, Labour's support was down by nearly five points on 2003.

The party lost control of councils, too. Having won control of 13 councils in 2003 - around half of which it might still have won under STV - Labour was left this time with just two, Glasgow and North Lanarkshire. No other mainland council now has one group in overall control.

How, though, does STV itself emerge from last Thursday? Were voters able to cope with the new method of voting? Just how proportional did it prove to be? And did voters use the opportunity provided by the new system to vote for individual candidates they liked rather than just the party they preferred?

Voters had significantly less difficulty completing the STV ballot paper than they did the redesigned ballot paper for the Holyrood election. In the latter, 4 per cent of the constituency votes and a little under 3 per cent of the regional votes were ruled invalid. The final tally for the local elections looks to be closer to 2 per cent. Among the eight councils we have examined in detail, the highest incidence of invalid votes was in Renfrewshire, at 2.5 per cent, while in Edinburgh it was 1.2 per cent. These figures are very much in line with the equivalent figures for local elections in Northern Ireland.

Most voters also seem to have realised they could cast more than one vote. Although voters appear to have expressed more preferences the greater the number of candidates on the ballot paper, the detailed figures released by some councils suggest the vast majority expressed at least two preferences and where there was a reasonably lengthy ballot then at least three was the norm.

But how proportional were the results? Our graphic compares the proportion of first-preference votes won by each party with their share of the seats in eight councils. For the most part the outcome in seats appears to have been reasonably proportional, after bearing in mind that wards of just three and four members
are too small to be perfectly proportional. Larger parties tended to secure somewhat more than their proportionate share, smaller parties rather less, but they were not squeezed out entirely.

In East Ayrshire, East Lothian and Renfrewshire, Labour won more first-preference votes than the SNP, but only won the same number of seats. However, STV is not just about first preferences. Liberal Democrat candidates in particular often did well in picking up second and lower preferences, and in both East Lothian and Renfrewshire this helped the party deny Labour a vital seat. In contrast, the Conservatives often
failed to pick up many lower preferences from other parties' supporters, ensuring they sometimes lost out in tight races for the last seat.

But did voters vote for individuals, not just parties? The popularity of individual candidates certainly mattered. In wards where a party put up more than one candidate, it was not uncommon for one of those candidates to receive far more first preferences than the others. For example, in Garscadden in Glasgow the outgoing provost, Liz Cameron, won more than 2,700 first preferences, while the former council leader, Jean McFadden, secured just 1,100.

Equally, voters offered more than one candidate from the same party did not feel compelled to choose all the candidates of that party before voting for a candidate of another party. In Garscadden, only 42 per cent of Liz Cameron's voters put Jean McFadden as their second preference. And, while this was exceptional, it
was not uncommon for only around two-thirds to three-quarters of a party's first-preference
voters to stick to the party ticket.

On the other hand, 39 fewer Independent councillors were elected. Most - 22 - were lost in Highland, where the political parties used the greater opportunity provided by STV to challenge Independents. While they rarely displaced all of them, at least one party candidate secured election in all but one Highland ward. Outside rural Scotland, however, Independent councillors prospered - eight more were elected in central-belt councils where party politics has its strongest grip.

That personal popularity matters under STV will doubtless not be lost on Scotland's new councillors as they consider how best to secure re-election in 2011.

John Curtice is a professor of politics at Strathclyde University.