Abstract
A new institutional mechanism is proposed for democratising politically difficult changes in tax-spending margins. Samples of the population could be asked to decide how much more (or less) they are prepared to pay in return for increased public spending in certain areas, such as health and education. Information will be supplied to them indicating how the corresponding tax increase (or decrease) will fall on different sections of the population. This process will occur on an annual basis, perhaps around two months prior to the budget, and will be used in an advisory fashion. The process has major advantages both over the current arrangements and over standard hypothecation.

The problem to be addressed

Political pressures over tax have imposed very great constraints on the parties and on debate. The fear is that the electorate, despite their expressed views in social attitude surveys, are not prepared to pay more tax yet at the same time continue to demand more spending. This unsustainable position puts political parties into an apparently impossible position and risks creating an unrealistic political debate, a massive budget deficit and widespread dissatisfaction with politicians. The administration in power becomes the target and focus for the dissatisfaction of all groups involved: the public sector unions wanting more pay, consumers wanting better services and taxpayers wanting lower bills.

One proposed solution has been to move towards forms of hypothecation. This would involve clearly separated budgets for particular forms of tax-spending (such as pensions, health etc.). However, there are major problems with most currently proposed forms of hypothecation. They tend to bring rigidity, so that money cannot be transferred between budgets: this undermines governments abilities to develop cross-departmental strategic policy and also tends to discourage big picture political debate. Second, hypothecation per se does not necessarily overcome the political problem of tax-spending decisions being seen as taken primarily by politicians and government, for example, many - perhaps most - contributions would continue to be compulsory and set by government even if they become ring-fenced for a particular form of spending.

The solution: a new institutional mechanism for consultation?

The proposed alternative is to explore ways of consulting that allow the electorate choices but that at the same time require the taking of responsibility for those choices - opportunity with responsibility in tax-spending decisions.

The proposal consists of the following:

· A representative sample of the population is selected at random to engage in the consultation exercise. This should require no more than an hour of each respondents time, and co-operation should be required in the same way that it currently is for jury service. The constraints on sample size follow from: desired confidence intervals of population estimates, resistance of sample to direct manipulation (e.g.: bribes), and cost. Given the reliability and common use of population samples for other purposes, it is ridiculous that the only forms of consultation currently used in the political process are, on the one hand, general elections and the occasional referendum, and on the other, devices such as Commissions of the great and good, obscure committees and poor quality double-guessing via the media. A sample of 2,000 to 5,000 drawn at random from the electoral register would probably suffice.

· Respondents would be required to answer a short questionnaire. For each of a series of high familiarity spending areas, such as health and education, the subject would be asked: (a) working on the assumption that any change in spending will result in a corresponding change in tax, would you like to see a change in current levels of service provision in [health, etc.] and (b) what level of change would you like to see. Responses would be given in the form of a real terms percentage change (e.g.: -2%, -1%, -0.5%, no change, +0.5%, +1%, +2% etc.) Alongside the question would be a grid indicating, for each level of marginal change, an indicative example of what change in tax this would lead to for people in different socio-economic circumstances, so that the subject could estimate the effect on him or herself and family. In addition, information would be given on what the extra spending (or cut) could be used to pay for (or would have to be cut).

It would also be possible to gather information on the respondents own socio-economic circumstances and use of services for the later analysis (see below).

· The consultation process would occur every year prior to the budget - perhaps around late August or early September. Results of the survey would be publicly available. It would become a regular event and one which would be expected to become a major focus of attention for many pressure groups including the public sector unions (and the affected private sector), consumer groups (patients, parents etc.) and other interests. The findings of the consultation process would not be formally binding, in that the government will consider them alongside other factors, but the findings would be expected to carry considerable influence over marginal adjustments, especially where the results are very clear.

Advantages

· The process creates the political space within which the electorate can decide to spend more (or less) on a given form of spending - something that political parties have become wary of risking, especially concerning possible increases in tax-spending. In so doing, the process will foster a more constructive and realistic political debate about tax-spending issues.

· It represents a real example of democratisation, giving the electorate both an opportunity to exercise power and a responsibility for the consequences of their decisions. It has the side-effect of no longer making it possible to blame the politicians for everything.

· The process will transform the relationship between the public sector unions and the government, notably over pay. Rather than focusing exclusively on the government in power, public sector interest groups will increasingly have to focus their attentions directly on the public and tax-payers they serve. Although the government retain discretion about tax, spending and pay, it is clear that, for example, if the public sees a section of the public sector receiving a series of significant pay increases without any improvement in services, then they will become increasingly hostile to any further increase in tax or spending in that area. If teachers or doctors want pay increases above the rate of inflation, they will have to justify them with the public directly. Of course, in the event of a strong vote against tax-spending increases in a given service(s), for example corresponding with an economic recession, then this will be very important ammunition for government to use in its negotiations with the public sector groups involved (especially over pay).

· The consultation process is non-binding to the government and considerably more flexible than formal hypothecation. Hence options still exist such as transferring money between budgets or sectors for strategic reasons, overall levels of spending can still be adjusted according to wider or longer-term constraints (such as economic performance), and decisions about how money is spent inside of budgets (priorities) still remain under the control of existing arms of state (though consultation processes could be extended into this area).

Problems?

There are a number of potential problems and criticisms of the mechanism that are worth considering, though none seem very serious:

· It can be argued that an individuals decision about a marginal change in tax-spending will depend on an estimate of overall tax-spending levels. Hence, an individual might be happy to pay for an extra 2% of spending on education today, but might not be prepared to do so if this increase coincided with an overall increase in 5% resulting from the sudden need to reduce government borrowing. This implies that the system is likely to need fairly stable overall conditions to work meaningfully. However, this seems a reasonable objective itself.

· The mechanism implies a loss of power for central government - perhaps even a loss of sovereignty. However, since the loss really constitutes a transfer of power back to the people, and given that it is limited to a consultation process only, this seems a weak critique. Furthermore, given that it can be argued that the politics of the present debate effectively foreclose many tax-spending decisions, it could be that the mechanism will actually prove empowering in that it will give governments back the authority and confidence to pursue certain options.

· What happens if the sample vote for unexpectedly substantial increases in tax-spending? This only seems to create an opportunity. It remains for the government of the day to weigh in all other factors before making final decisions, (though it should at least explain why it has reached these decisions).

· What happens if the sample vote for unexpected cuts in tax-expenditure? Again, the government is still in the position of making the final decision. It might be that there are other considerations to be taken into account and budgets that can be borrowed from such that spending could or should be protected. At least the government would be making the decision with its eyes open to the wishes of the electorate.

· Is there a danger that, as is sometimes said of referendums, views expressed will be as much an expression of attitudes towards the government than of a real view? But, unless the government has expressed a clear view about which way the population should vote, it is difficult to see how this punishment would arise. The only serious punishment effect would seem to be one indicating that the electorate had little faith in the relevant public service or the way that the government was managing its spending. Is this such a bad thing?

· There are statistical and sampling issues to be considered, most notably concerning the geographical area of the sample. For example, it could be that an area with a very elderly demographic profile might be more in favour of spending on health but less in favour of spending on education. This might suggest that regional polling, in the event of significant regional differences in such response-sensitive variables, could create problems and that national polls are best (as suggested above). However, national polls would be insensitive to regional differences in the performance and quality of services implying that other mechanisms must be applied to respond to such variations.

· People might be suspicious of how representative a representative sample is. Many people have a poor intuitive grasp of sampling theory, and this might undermine the credibility of the mechanism. However, this simply suggests that (a) public education should occur and (b) the process should be gradually introduced over a number of years, via a gradual increase in the weight given to its results.

· There is a limitation of the process in that there are certain areas of spending, especially those which concern spending types that have little tangible impact on most of the population most of the time, that are unlikely to be suitable for this type of consultation process. It is difficult for the average individual to work out the value to them of some marginal change in tax-spending designed to protect the population from some very low risk occurrence, e.g.: on defence spending, obscure research, etc. Also, some forms of spending cannot be neatly segmented. These factors may limit the applicability of the process to the more familiar and ubiquitous forms of state activity.

Conclusion

The tax-spending consultation process outlined in this discussion paper has major political and procedural advantages over both the present arrangements and standard hypothecation. It should be of particular interest to an incoming Labour administration because:
(a) it offers the potential for a creating a greater range of tax-spending options (notably increases) without political or electoral fall-out;
(b) it provides a refreshing example of giving power, opportunity and responsibility back to the people and at little cost either in terms of money or legislation
(c) the mechanism has the potential to significantly alter the terms and form of political debate in the UK, both in terms of the apparently paradoxical stand of an electorate wanting higher services and lower taxes and in terms of the focus of various pressure groups, not least being the public sector unions.