The idea of referenda on local authority budgets must be examined again. Superficially, it looks attractive, as do most proposals for referenda, until they are investigated. It is proposed as abolishing capping and giving local authorities freedom from central control over expenditure, provided they can convince their own local electors. But it retains the fundamental objection to capping. The Government would still set a maximum level of expenditure for individual authorities, as in capping, only now proposals above that level would be used as the trigger for a referendum, on expenditure the level of council tax. 

The Government's labelling of any expenditure above a certain level as excessive undermines the principle that local authorities should make their own decisions on their own expenditure and level of taxation. An authority spending above that level should not be regarded as over-spending; it merely has different views from central government. It would be surprising if all authorities had the same views as central government. The Conservative party has not made clear, any more than the Government, why it is necessary for central government to have a view on how much each local authority should spend. The economic case has never been made. The political case should never be made by any government which says it wishes to free local government from central control, and all the main parties say just that.

The fundamental objection to the Conservative party's proposal is that it will damage our system of government based on representative democracy. The challenge for democratic renewal is how to strengthen that representative democracy. The onus on the advocates of referenda is to show how they would enhance and not undermine representative democracy. The Conservatives need to set out criteria that can indicate when referenda are appropriate. Even the most fervent advocate of referenda has to admit that most decisions in central and local government cannot, in practice, be made by referenda. Mr Cameron needs to demonstrate why referenda are appropriate for budgets and for what decisions they are not appropriate. 

We believe referenda should be used only on issues that will produce a clear decision in favour of one alternative. They are not appropriate for decisions which involve balancing factors against each other and cannot be reduced to a simple yes or no vote. They are not appropriate for issues that are inter-connected with others, nor for issues which have been at the heart of local or general elections on which the public has already made a decision. While a council's final decision on the budget may appear to be a yes or no, that stage has been reached only after a prolonged period of decision-making, as the pattern of expenditure is built up. 

Budgetary decisions are interconnected with many other decisions that cannot be taken into account in a one-off vote. A referendum decision could undermine the local election. The party elected could be committed to increases in expenditure, which could later be voted down in a referendum. The council, as the representative body, would be weakened by that decision and the value of the electoral process would be thrown into doubt.

Other considerations need to be taken into account. The referendum will be biased from the outset, since the Government will have labelled the budgetary increase as excessive, bringing its weight to bear on the electorate and undermining the local elected council. The form of the referendum is biased. It will focus attention on the increase in council tax. Against that has to be weighed the more complex impact on services of any reduction in the proposed expenditure, which will remain uncertain to the electorate at large. 

No tax is popular, but features of council tax make it particularly unpopular. It is widely seen as unfair. In addition, because the tax base is not inflation-proof, the rate at which it is set has to be increased each year to keep pace with inflation. Council tax is paid directly by the taxpayer, unlike most other taxes that are paid indirectly. Compared with council tax, they are stealth taxes. The electorate are likely to be hostile to the proposed increase in council tax, as they have not gone through the same process as councillors when preparing the budget, weighing the burden of increasing council tax against the implications for expenditure on services.

The council can put its case, but the focus on the increase of an unpopular tax, coupled with the Government's statement that the tax is excessive, makes it unlikely that councils will risk a probable defeat. One would still have capping because most councils would cap themselves rather than risk being forced to call a referendum, even if they believed greater expenditure was needed. There can be no surer way of weakening local representative democracy. 

George Jones is emeritus professor of government at the LSE, and John Stewart is emeritus professor at INLOGOV