As an exercise in participative democracy, it seems badly flawed. Brazilians will vote in a mandatory referendum tomorrow to decide the question, "Should the sale of firearms and ammunition be prohibited in Brazil?"
There should be little doubt about the answer. Gunshot wounds kill more than 107 Brazilians every day, more than traffic accidents and fewer only than heart and brain disorders.
Restrictions on the owning and bearing of firearms were introduced in December 2003 with striking results. In São Paulo, one of Brazil's most violent cities, the number of people killed by gunshot wounds fell during the following year by 22 per cent, resulting in 2,630 fewer deaths. Disarmament became a popular cause. A call on the population to hand in their weapons resulted in 385,800 guns being surrendered by July this year. In opinion polls before campaigning on the referendum began, 81 per cent said they were in favour of the ban.
Yet since then something has changed, and the latest polls suggest that prohibition of the sale of firearms may be rejected.
The trouble began with the 2003 law. It called for a referendum not because participative democracy was held to be more appropriate in this case than the usual representative kind but as the result of congressional horse-trading.
The law, known as the statute of disarmament, prohibited ordinary citizens from carrying weapons and made buying them more difficult. It had wide support among legislators but also faced stiff opposition. "The gun lobby in Congress is small but efficient," says Denis Mizne, a security expert and director of an anti-gun group called Sou da Paz ("I'm for Peace").
In exchange for letting the bill through, the gun lobby insisted its proposed ban on selling guns and ammunition be put to a referendum.
The question posed is confusing. Those who are against guns must vote Yes (doubly counter-intuitively, Yes is number two on the electronic voting machines) and those for them, No (number one). A daily newspaper, O Estado de S. Paulo, recently put the question to 100 people in the centre of São Paulo: 29 of them got it wrong (that is, they voted against the ban when they were in favour of it, and vice versa).
More damage to the disarmament cause has been done by the campaign to support it. Supporters have made questionable claims in a questionable manner, pitching "us" against "them" and exaggerating the rights of rural communities to own guns so as not to alienate them. It was forced by the courts to remove elements of its advertising, to the delight of the No campaign.
Instead of drawing on the evidence, the Yes campaign initially relied on singers, actors and models issuing vague exhortations along the lines of "give peace a chance". That has alienated many potential supporters. Raimundo Fagner, an ageing pop star, struck a chord when he said: "Whenever a load of artists start talking, you know there's something wrong. They're a bunch of sheep who say what they're told to and don't think about anything."
The No campaign has done a better job. The ban on selling guns, it says, will disarm ordinary people and leave criminals with as many guns as before.
Late in the day, the Yes campaign has begun to tackle this head on. Less than two weeks before the poll it changed advertising agencies and adopted a harder tone.
It has powerful arguments on its side. Very few gun deaths - just 5 per cent - take place during robberies. Most follow arguments in bars or in the home between friends and family. Legally bought guns are frequently stolen (40,000 in 2003), supplying the illegal market. A study released this week by Rio de Janeiro police shows that 61 per cent of guns used by criminals were bought legally. Two children are admitted to hospital every day with gunshot wounds, most the result of accidents.
The Yes campaign has regained some support and the latest polls show an even split. But whatever the outcome tomorrow, the referendum could depend more on chance and marketing than on reasoned decision-making.
