Grey areas surrounding the green paper on electoral reform could allow the government to 'repeat their tricks' and manipulate public opinion during the public consultation, a top polling academic has warned. University of Hong Kong pollster Robert Chung Ting-yiu said 'very many alarm bells have been ringing' since the green paper was released. These included the short consultation period of only three months and the 'arbitrary standard' requiring the final proposal to have public support of 60 per cent. Most crucially, he said, the government had yet to declare how it would evaluate and collate the submissions received, meaning that no matter how many signatures are obtained in support of a particular proposition, they could all be discounted as being of 'low quality'. 'At the end of the day, if you want to manipulate the results it's not difficult,' Dr Chung said. A spokesman for the Constitutional and Mainland Affairs Bureau said two elements would play a critical role in the government's assessment of reform options, in addition to the written submissions. One was opinion polls conducted by universities and think-tanks indicating whether there was 60 per cent support for any proposal. The other was whether such a proposal was likely to gain the support of two-thirds of the Legislative Council. In a 2003 paper, Dr Chung argued that the colonial and post-colonial governments had engineered the results of two public consultations - on direct elections in 1987 and on the Article 23 national security legislation in 2002. In both cases, the government did not explain how it would analyse the responses, he said in an interview explaining the paper's conclusions. 'At some point when they compiled the report, they chose a system which gave support to the government proposition. Basically, they manipulated the concept of 'submissions',' he said. 'The British government concluded in 1987 that the public was opposed to direct elections, and the Hong Kong government in 2002 concluded the public was in favour of legislating a security bill. In both cases, the government engineered the result of the analysis and lied with statistics,' Dr Chung said. The paper said that in both cases the government collated public responses into different categories, rather than simply count how many people supported a certain proposition. Petitions were given less value because they were 'products of mobilisation', the paper said. But 'standard letters' - the method adopted by pro-government groups - were included in the higher class of submissions. 'I could see no difference in the nature of someone signing a petition form someone signing a standard letter. They should have been treated in the same category,' Dr Chung said. He said the democratic camp had shown itself to be wiser this time by issuing 'recommended answers' because these would be treated as 'standard letters' and would therefore be accorded higher value. Secretary for Constitutional and Mainland Affairs Stephen Lam Sui-lung last week ridiculed the 'recommended answers' as possibly at odds with 'democratic principles of free choice'. Dr Chung acknowledged that the 'quality of opinion' of such submissions might not be as high as some original answers. 'But if the government is not going to set down the rules of the game right from the beginning, they have no right to criticise anything. What else do you expect them to do if you don't set down the rules?' Dr Chung said.