According to the Bible, "The poor will ye always have with you." The Bible need only be right if poverty is seen as relative deprivation. In that case some people are always bound to be poorer than others. The search for a perfectly equal distribution - apart from being full of conceptual ambiguities - is in practice likely to lead to hell on earth, with much bigger disparities of power in exchange for relatively modest (if any) improvement in the relative possession of the worse off.

On the other hand it does not quite work to define poverty only in absolute terms. It is of course worth knowing how many people there are who are hungry, or homeless or without clothes or heating or lighting. There is no need for anyone to be involuntarily in such a state in a moderately affluent western society.

But on analysis this simplicity disappears. What do you do about people who would have enough to eat and could afford to do so, if their diet were planned by nutritionists, but in practice do not have enough to do so? Do we therefore want to extend this principle to include anyone of any income level who does not or cannot spend his or her income wisely?

The problems go further in an advanced industrial society. The kinds of cheap consumption available to a Bangladeshi are not available to us. An Englishman does not have the option of paying lower train fares and sitting on hard wooden benches. The very basic foodstuffs available in third world retail outlets are difficult to find in a western metropolis.

One can go even further. There are such things as conventional necessities. A citizen who cannot afford a television set is in some sense deprived of a common point of reference with fellow citizens - of which he or she would not have been aware, living on the same income in 1850. But a line has to be drawn somewhere. To say that someone who cannot afford a holiday in the West Indies, or many meals at expensive restaurants, is poor is simply playing with words.

There seems to me no escape from the notion that poverty is in part a relative and in part an absolute concept. If you put all the emphasis on the relative part, then the main effect is to spread misery. For it means that however rich our societies become and however far-reaching the applications of science, social critics will always be able to complain of poverty; and, to the extent that their message is diffused, people will feel worse off than they need feel.

At root of much of the argument is the unanswerable question whether the riches of the wealthy or the more moderate affluence of the middle income groups is the cause of the poverty of others?. If someone makes or applies a new discovery, like Henry Ford, which benefits everyone but becomes rich from it himself, he is not causing poverty. On a reasonable use of language he is alleviating it.

But for the mass of upper and middle-income citizens with no such talents or opportunities, the question is much harder to answer. If there is some redistribution which could be undertaken, along Rawlsian lines, which really would make the poorest better off, it should be undertaken. But is the failure to make it a cause of poverty? We are here up against an ambiguity in the word "cause", which we find in many other key concepts in the social and even natural sciences.

The acid test of whether someone who campaigns against poverty or deprivation is activated by malice against the better off or concern for the poor is the following:- how does he or she react to an improvement in the income or wealth of the better off which has no adverse effect on those at the bottom or in the middle?

Anyone who thinks such an improvement impossible should consider the following. A number of people have become rich through setting up internet companies. They have not become so by grinding the face of the poor; they have done so by venturing into a new form of activity which increases total national income, even if by not as much the enthusiasts suppose. They have done nothing to harm the poor and may indirectly have benefited them, even though on some statistical measures the "degree of inequality" has increased.

It so happens that, whether intentionally or by happy accident, the Rowntree definition of poverty passes the anti-resentment test with flying colours. On that definition, a poor household is one whose income is below half the median - the median is simply the mid point, where there are as many households above as below it. The arrival of the internet millionaire has not increased poverty on this view - the changes have all been at the top end of the income distribution. But on the government view of distance from the average it almost certainly has.

The Rowntree definition looks superficially similar to another definition of poverty as below half average income. But there is all the difference in the world between them and the difference should not be regarded as merely technical.

This is brought out very well in some of the tables near the beginning of the Rowntree survey "Monitoring Poverty and Social Exclusions, 1999".

They show that a reduction in the basic rate of income tax appears to increase poverty on the "distance from average" definition. Even an increase in the work incomes obtained by the better off half of citizens, without any government action, appears to increase poverty.

The reason is that anything that happens which benefits the better off sections of the population, or those in the middle, will increase average income. If the bottom quarter remains stable, then their difference from the average will increase, even if nothing has happened to increase poverty on a common sense view.

I have, however, mixed feelings about all the detail about what the poor cannot afford given by surveys such as Rowntree. There is of course a strong case for spelling out poverty in concrete terms. In the past it was easier to make out a case for redistribution by pointing to the number of households without inside lavatories or bathrooms than by abstract statistics of income distribution. And yet is it really so surprising or so disquieting that better off people are healthier than poorer ones? There is more to health than medical care; and it would be both sad and astonishing if people on higher incomes spent all their extra resources on conspicuous consumption and overlooked the needs of their own person.

The figure that impressed me most in the Rowntree study was that it would be possible to abolish poverty on the Rowntree definition by a relatively modest expenditure of public funds - some £6bn to £10bn or 1 per cent of GDP. But the authors then go on to spoil their case by saying that the difficulty is to do this while diminishing reliance on means tests. I have long advocated the goal of a basic income (not minimum wage) adequate for a conventional subsistence or all. But this is not round the corner; and the elimination of poverty in the Rowntree sense will only be delayed if means testing is written off as inevitably carrying a stigma - the income tax has always been based on a means test.

Why moreover put all the emphasis on increases in income tax to fulfill this goal? The basic rate of tax is, as the Institute for Fiscal Studies reminds us, one of the less important statistics in the tax system? But harping on it only increases popular opposition to redistribution, even if it gives a puritanical thrill to some. Why not talk in terms of increased VAT rates or petrol duties and the like?

I leave for last the doubts I have about the fashionable New Labour term "social exclusion". This adds nothing to what is conveyed by the simple word poverty. Of course we can use words to mean what we like; but they have connotations which affect how they are received. Exclusion suggests that it is the action of other people who make some citizens worse off. Yet if someone is not able to drive a car, what is gained by saying that society excludes him, with its ridiculous implication that his car-driving friends should feel guilty on his account?

A black South African who was not allowed into an expensive restaurant because of apartheid was excluded in a meaningful way If he cannot afford to go to the same restaurant today it is more accurate to say that he is poor without trying to make others feel guilty. As usual there are intermediate cases. A Jewish businessman who was not allowed into some prestigious golf club could say that he was suffering from social exclusion. But if there were plenty of other clubs he could join, he was only playing up to the anti-Semites by showing his resentment. He might have done better to wait until he could buy the club. Similarly for an Asian today.

What do we say about the opinion poll data which suggest that personal happiness is moderately correlated with relative income, but not with absolute income?. In other words rich people are a bit happier than middle income people; and the latter a bit happier than low income people; but as the average or typical income level rises the population as a whole do not feel any happier.

In my view, nothing very much. Envy and jealousy are facts of human nature and there have been interesting speculations about their evolutionary role. But it is one thing to accept them and another thing to use public policy to reinforce them and to make a vain attempt to satisfy them.

My bottom line is that the poor lack money - that is income and capital. I would hope that more resources could be transferred to them by a reasonably generous electorate without guile or playing on guilt feelings. It would be even better if, by means of education and example, those who are now poor, or their children or descendants, could acquire more original income so that such transfers are less necessary. But do not let us fool ourselves about how much education can achieve. There will always be people - including some intelligent and sensitive persons - who will lack the particular skills which are in demand in society at any one time. To try to force them to perform according to some ideal yardstick is ultimately a threat to liberty (their liberty) which some of us value more than equality.