Starting late Sunday night, results from the constituency vote started to appear on television. Many voters had taken the "no-vote" option, and surprisingly many had spoiled their ballot papers.

In the previous election last year, spoiled ballots comprised 2.9 per cent of all votes cast for the party list and 6 per cent of the constituency vote. This time, the figure for the constituency vote was more than double at around 13 per cent. Many ballot papers were defaced with slogans and abuse directed against Thaksin. Many were probably left blank because people thought that "no vote" meant not marking the ballot at all. Lots of these spoiled ballots were clearly protest votes, but we cannot tell how many.

Most of the television channels and newspapers did not report details on the "no-vote" tally and spoiled ballots, but Channel 11 displayed the full data, and Matichon newspaper printed a fairly complete set for 397 of the 400 constituencies. Adding these data up showed that about 52 per cent of all votes nationwide went to Thai Rak Thai, 2 per cent to other parties, 33 per cent to the "no-vote" option and 13 per cent were spoiled.

Thai Rak Thai got 15.4 million of this constituency vote. Where did the other 800,000 votes for the party list come from?

In last year's election, Thai Rak Thai gained 2.4 million more votes on the party-list vote than on the constituency vote. Some voters used their constituency vote to choose a local candidate they liked but then used their party list vote to select Thai Rak Thai. That behaviour makes sense.

But in this year's contest, the choices facing the elector on the constituency vote and the party-list vote were not really any different. The minor parties were not important. On both polls, the voters' real choice was between Thai Rak Thai on the one hand, or a protest vote ("no vote" or spoiled ballot) on the other.

But the percentage distribution of votes between Thai Rak Thai, minor parties, abstentions and spoiled ballots was very different for the two different types of ballots. On the constituency vote it was 52-2-33-13, and on the party list it was 56-7-31-6.

The most striking difference is that 3.8 million people spoiled their ballots for the constituency vote but only 1.6 million for the party-list vote. In other words, at least 2.2 million people spoiled one ballot but not the other. In addition, at least 600,000 decided to cast a "no vote" on one ballot but not on the other.

It's very difficult to understand why the number of protest votes ("no vote" or spoiled ballot) was around three million higher on the constituency vote than on the party-list vote. You'd expect the opposite. If you wanted to register a protest against Thaksin, it would make more sense to do it on the party-list ballot. On top of that, more people voted for minor parties in the party-list vote than on the constituency vote. Again, you'd expect the opposite.

No details have come out regarding the party-list returns except for Bangkok. Comparing the constituency vote and the party-list vote produces some odd results. In Bangkok constituency 15 for example, more than one in five of all voters made a different choice on the two ballot slips. They apparently went into the polling booth and within the space of a few seconds marked one ballot slip in one way and the other in another way. It's difficult to imagine why.

All of the Bangkok constituencies are similarly wayward. There is a general trend of fewer spoiled ballots on the party-list vote compared to the constituency vote. But in other respects, the differences seem almost random. Sometimes Thai Rak Thai scored higher on the party-list than on the constituency vote. Sometimes the "no vote" was higher. Sometimes votes for the minor parties are higher. It's very difficult to understand what was going on. Perhaps these early figures are wrong. Perhaps they are a clumsy effort to reach a target.

On the afternoon of April 2, Thaksin was televised at Chiang Mai Airport. Full of confidence, he insisted that all parties would have to abide by the result of the election because it was the will of the people.

He called a press conference in Bangkok that evening. It was rumoured he would claim victory. At the last moment, the press conference was cancelled. We know that Thai Rak Thai had conducted an exit poll, and presumably it would have shown something close to the 52-2-33-13 result on the constituency vote.

When the television channels started their election coverage early the next day, there were some surprising features. None of the free-to-air channels made any attempt to add up constituency-by-constituency returns to give estimates of regional or national patterns. No returns from the party list were reported at all. The website of the Election Commission (EC) was designed to hide information about the "no vote" and spoiled ballots.

Later in the day it was announced that Thaksin would appear on television that evening. The television reporting of the election tailed off. Channel 11 stopped displaying the results and showed monks chanting in the Temple of the Emerald Buddha.

That evening Thaksin claimed victory with 16.2 million votes on the party list. He said the data came from the Interior Ministry, which was strange because that ministry was not involved in running the election. The 16.2-million figure represented 60 per cent of the valid votes cast. By coincidence, the equivalent figure in last year's poll was almost exactly the same. The EC has since adjusted the figure even higher, to 16.4 million.

Such a strange election.

Chang noi