Interviewing Sister Connie Driscoll on the radio during her visit to New Zealand in March 1997, Brian Edwards expressed the opinion that charity was 'offensive'. I was gobsmacked by his attitude and have been pondering it ever since. Sister Connie is a person who is as close to a saint as I expect to meet in this life, and the story of her work for drug-addicted welfare mothers in Chicago inspired her New Zealand audiences, including ACT's annual conference. What kind of person, I wondered, could regard such charity as 'offensive'. Not, of course, that Brian Edwards was against the provision of welfare, but he thought it was the responsibility of government, rather than voluntary charities, to provide it.

It is doubly puzzling to discover such an attitude towards charity when compassion is often elevated to the highest of modern virtues. Is there not something ironic here that needs to be explained? Since compassion – literally, 'suffering with' someone – would seem to lead so naturally to spontaneous charitable action to relieve the suffering, how does the one come to be praised and the other disparaged?

Could the answer be that charity is indeed, as Brian Edwards claimed, 'capricious' and 'unreliable', and that state welfare does the job better? This came to be believed around the turn of the century, when New Zealand and Australia led the world in imitating the state-run social insurance policies that had been pioneered in Germany in the 1880s under Bismarck. The continuous expansion of the welfare state since then has reflected the widespread conviction that welfare is above all a state responsibility, and nowadays is the main justification for high levels of taxation. Whereas in previous centuries the state played a minor role and left most welfare delivery to voluntary mechanisms like families and charity, in the twentieth century that relationship has become inverted: the charities have been relegated to a minor role, and even that role is largely supervised and subsidised by the state.

So residual has the voluntary sector become in the way we think about welfare in New Zealand that no reliable statistics about it exist. According to Jim Datson, the immediate past president of the Fundraising Institute of New Zealand, the National System of Accounts puts the value of the voluntary sector in 1993 for GDP purposes at $771 million, but the New Zealand Association of Philanthropic Trusts estimates its true value at around $2 billion. The Inland Revenue Department calculates that 'gifted income' amounted to $259 million in 1993. Census figures suggest that New Zealanders volunteer around 500,000 hours of work a year to the voluntary sector. But these figures are very broad-brush, and the term 'voluntary sector' includes many associations, like sports clubs and political parties, that are not charities. The major charities could consider commissioning a proper statistical database of the voluntary welfare sector as a way of raising their profile. Meanwhile we should bear in mind that the charities' incomes are derived not only from gifts but also from government subsidies worth about $580m a year, commercial activities, and accumulated resources.

More accurate information about the voluntary welfare sector is especially needed as confidence in the welfare state declines. In New Zealand, the share of working-age people dependent on state welfare has doubled since 1985 and now stands at one-fifth. Over half of all working-age welfare-state beneficiaries have been receiving their current benefits for more than a year. This rise in welfare dependency continued despite the rise in employment and fall in unemployment in the mid-1990s. Recent publications by the Business Roundtable have argued that state welfare has created perverse incentives to become dependent on welfare – to become 'pauperised', in the unsqueamish terminology of previous centuries – and that voluntary associations are more likely to dispense welfare in a way that encourages its beneficiaries to become self-supporting. What is needed is a new division of labour between the state and the wider civil society: the state's comparative advantage is in providing a safety net of last resort, while income security and general personal welfare are in normal circumstances best safeguarded by mechanisms such as private insurance and savings, by family responsibility and by the voluntary sector. The effectiveness of all these mechanisms would be improved if disposable incomes were increased by tax reductions.

In this address I don't wish to rehearse these arguments, beyond commending to you the publications in which they are set out. What I want to do here is to explore Brian Edwards's claim that charity is 'offensive' and the implications of that claim. I believe that the resistance to modern arguments for welfare reform stems largely from a conviction that charity undermines true compassion with unworthy sentiments: that it induces a patronising contempt in the donor, and puts the recipient in a position of humiliating dependency. The welfare state, in contrast, is believed to avoid these effects because it somehow confirms the bonds between donor and recipient as members of the wider community, and distributes benefits without drawing a moral distinction between the 'deserving' and the 'undeserving'.

The most compelling and influential attempt to characterise state welfare as the embodiment of compassion was made by Richard Titmuss, a British academic at the London School of Economics whose work provided one of the intellectual foundations of the postwar expansion of Britain's welfare state programmes. I shall refer to Titmuss several times, not just because of his profound influence, but because the flaws in his arguments are so instructive. His famous 1971 book The Gift Relationship grew out of a debate in the 1960s about voluntary donations of blood in Britain. Hospital stocks of blood were always in short supply, and some policy makers were advocating a market in blood, such as existed in several other countries, to ensure that the supply of blood for medical purposes matched the demand for it. Titmuss strenuously opposed this proposal, on the grounds that it would replace the 'gift relationship' that existed between voluntary donors and recipients with the 'exchange relationship' of the marketplace. To the objection that a market in blood might save lives, Titmuss replied that altruistic behaviour was imitative and, if encouraged, would spread by force of example and ensure that sufficient supplies of blood were forthcoming.

But why exactly was the 'gift relationship' so morally preferable to the exchange relationship? Titmuss valued giving blood because it was, he said, a perfectly altruistic unilateral transfer to a stranger, motivated neither by expectation of reciprocal reward nor by favouritism towards its unknown recipient. The impersonal nature and anonymity of the gift relationship between donor and recipient relieved them of feelings of moral superiority and moral inferiority respectively. For Titmuss, this is what the welfare state was all about: it was a moral project because it bound citizens together as moral equals in the gift relationship. It realised the old socialist vision in which people worked not for personal gain but for the collective good: "from each according to his abilities, to each according to his needs".

This argument for the welfare state is what commentators have in mind when they say that the level of welfare state spending is a measure of a society's compassion. It sits rather awkwardly with another popular argument for the welfare state, namely, that we all have entitlements to certain welfare services. For if we have rights to welfare, these must be satisfied as a matter of strict obligation, not altruism. As well, some modern communitarians would object to Titmuss's stress on the anonymity and impersonal nature of the welfare state, since they regard these as undesirable features of the marketplace. Still, I think Titmuss was right to identify altruism as one of the motives underpinning state welfare, since there is widespread support for the view that the state should ensure that no one falls below a certain minimum of welfare. But this is not to say that other motives are absent from the welfare state, or that there is popular support for keeping people in quasi-permanent dependency as opposed to requiring of them some effort to improve their condition.

The real problem with Titmuss's analysis is that it takes absolutely no account of the actual outcomes of the welfare state. Instead of smoothly turning compassionate motives into intended outcomes, the welfare state has produced a state of affairs that no truly compassionate observer could desire. I have already referred to the growing problem of welfare dependency caused by state-administered programmes of unconditional poverty relief. Another problem is the diversion of a huge share of the welfare budget to the non-poor, especially in the form of universal services in health care, superannuation and education. Much of the additional $5 billion of public spending introduced by the coalition government is going to finance services from which middle-class people are the main beneficiaries. This can hardly be described as a manifestation of compassion: rather, it is the outcome of political processes that reward influential lobbies acting in their own interests and without any necessary regard to the public interest. As well, the growing welfare state has contributed to a weakening of the family and of individual responsibility.

In view of these disappointing outcomes of the welfare state, isn't it time we started to question what all this compassion talk is really about? There is the strongest contrast between modern public avowals of compassion and the spirit of practical concern that sustained the many vigorous voluntary welfare associations that flourished in the late nineteenth century and that Titmuss hoped would sustain the welfare state that has largely supplanted them. Very often nowadays, 'compassion' seems to consist of the expression of publicly virtuous sentiments in the media or the pulpit, but little in the way of genuine giving or personal commitment. Several critics of the welfare state have drawn attention to this stress on the feeling of compassion at the expense of concrete assistance for the object of the compassion. Robert Whelan, for example, writing in The Corrosion of Charity: From Moral Renewal to Contract Culture, gives this account of what he calls "compassion without tears":

In similar vein, Marvin Olasky, in The Tragedy of American Compassion, observes that 'compassion' has become a double-think term to mask the fact that the welfare state promotes personal meanness:

Indeed, the welfare state now embodies some of the features of charity that Brian Edwards appears to find offensive. At its worst, charity was an aristocratic pastime in which ladies bountiful derived a sense of moral superiority from dispensing amounts they could easily afford to the lower orders as way of keeping them content with their humble lot. Similarly, the welfare state has generated a class of dependants whom its intellectual supporters patronise by telling them they are 'victims of society' while arrogating to themselves the kudos of the system's presumed virtuousness. And just as the ladies bountiful resisted the political reforms that emancipated their serfs, so welfarist intellectuals resist the economic and social reforms that would restore to citizens control over their own incomes and responsibility for their own lives. In this context it is regrettable that many in the churches, whose intentions in this area are generally sincere and honourable, should have come to support so strongly the delegation of compassion to the state. This may be one of the reasons why their status as a serious moral force in society is so much diminished.

The truth is that the link between compassion and the welfare state, if it ever existed, has been largely shattered by perverse incentives and political opportunism. If we really want a truly caring society on the one hand, and a welfare system that works on the other, we must rethink from scratch the relationship between the two, and try to see where charity really fits in.

Recall the contrast that Titmuss drew between the gift relationship and the exchange relationship. Titmuss was assuming, as nearly everyone does, that gifts, being based on altruism, are morally superior to exchanges, which are allegedly based on self-interest. This simple-minded contrast is misleading and inaccurate in so many ways. In the first place, more welfare is generated by exchanges than by gifts. This is part of the general point that any wealth given away must first be produced. Both parties to a voluntary exchange benefit from it: it increases wealth. In a modern economy, the greater part of welfare is necessarily generated by spontaneous commercial exchanges among people whose labour is divided and specialised in their search for comparative advantage. Of course, gifts also promote welfare. In some areas, gifts may create as much welfare as do exchanges; it may well be possible, as Titmuss hoped, to secure a sufficient supply of medical blood by voluntary donations. But where it isn't possible, it is surely immoral to risk lives by prohibiting a market in blood in order to preserve the purity of the 'gift relationship'. When it comes to welfare, it's consequences that count, not motivation.

Again, the contrast between altruism and self-interest can be overdrawn since, in reality, the two motives are often mixed. First, there is a sense in which a pure gift benefits the giver as well as the receiver. This point was made by Shakespeare in The Merchant of Venice: in Portia's famous soliloquy on 'the quality of mercy', she says:

In a more pedestrian mode, an economist might notice the analogy between gifts and exchanges: just as both parties to a voluntary exchange gain from it, so a voluntary gift benefits the giver in the sense that he or she prefers to see the money, or whatever the gift consists of, in the recipient's hands rather than in their own. Yet this does not detract from the meaning of the gift; indeed, the spirit of giving is precisely one of wholehearted goodwill, and is spoiled if the gift is made reluctantly or resentfully.

This point brings into view yet another problem with Titmuss's ideal of a welfare state built on the gift relationship. In what sense can transfers financed by compulsory taxation be understood as 'gifts'? The idea is not wholly preposterous; I have already suggested that there is an altruistic consensus for a welfare safety net financed by taxation. But that would require a much smaller tax burden than our actual welfare state imposes. In practice, taxes are typically paid without goodwill, and are experienced as a loss rather than a gain. I say this knowing that some opinion polls purport to show that many New Zealanders are willing to pay more taxes to finance more welfare spending. The point is that their behaviour just doesn't bear this out. Who volunteers to pay more tax than they are legally obliged to? There is nothing to stop people sending additional cheques to the Inland Revenue Department or to the Department of Social Welfare, but how many people do? How often do voters reward a political party that promises to increase their taxes? I suggest these actions speak louder than words. Those who want to make society more compassionate are going to have to encourage more voluntary giving, which implies a cut in taxation and state welfare, not more of it.

The second way in which the altruistic and self-regarding motives are mixed is evident in long-term relationships in which welfare-enhancing exchanges occur which are not commercial ones but nevertheless benefit both parties. Consider ordinary friendships. Friendships are sustained by giving; if one party offered to pay for a gift, thus turning it into a commercial exchange, the friendship would be ended or damaged. But if over time the giving were all one way, that would usually threaten the friendship too. The long-term relationships on which we all depend are sustained by giving, but also by a principle of mutuality or reciprocity that ensures that both parties share the burdens and benefits.

Brian Edwards, like Richard Titmuss, appears to conceive direct personal charity as a wholly one-way relationship that is offensive because it places the receiver in a condition of humiliating dependency. But I suggest that this view overlooks the quiet but huge role that practical, reciprocal charity plays in our lives. In myriad ways, in all spheres of our lives, we are constantly giving and receiving aid in a mutual fashion. We understand that elementary cooperative, respectful and considerate behaviour is necessary if we are to realise our individual or collective goals. This sort of mutuality doesn't require any precise equality in the value of services exchanged as purely commercial exchanges might, but it does make us morally equal members of the wider communities in which we participate.

The principle of reciprocity is being invoked in the movement for welfare reform that is getting under way in several countries. In the United States, the welfare reform of 1996 obliges recipients of state welfare to undertake training or employment where they can. In Australia, the work-for-the-dole scheme requires young people receiving unemployment benefit to do the same. The community wage proposal has a similar motivation. The public is demanding that the people they are willing to help through their taxes make the effort to become self-supporting. Yet these state schemes are only imitating the way that voluntary associations have been operating for decades, even centuries. Marvin Olasky shows in The Tragedy of American Compassion that the voluntary associations that thrived before the rise of state welfare, and even after it, always made the avoidance of pauperism one of their highest priorities. As Sister Connie Driscoll explained to Brian Edwards, the homeless and destitute young women who enter her hostel have to agree to abide by certain standards of conduct and to help run the place, and can't expect just to be passively clothed, fed and sheltered.

One of the reasons why the voluntary sector is generally better placed than state bureaucracies to administer assistance in a way that helps recipients to recover their independence and self-respect is that it need not be inhibited about moral education. There is a case for believing that just as state education should not promote religious belief, so state welfare should not impose any moral code beyond elementary honesty and the minimal obligations of reciprocity that taxpayers may demand. But voluntary associations are surely entitled to try to impart the moral beliefs of their supporters to those they assist. Competition among the voluntary associations would reveal which codes of behaviour are most effective.

In an age dominated by an uneasy combination of two elite intellectual fashions, moral relativism and political correctness (and paradoxically, the two contradict one another), it's difficult to say in public, for example, that teenagers should not have children outside marriage, or that, if they do, that they should get married and stay together so as to give their child the best chance in life. But in a thriving voluntary sector, each association would have every incentive to broadcast the values it promoted, and the results obtained from them, as a way of attracting donations from the public. Meanwhile, those who wished to receive assistance would be able to exercise a certain degree of choice over the terms and conditions that came with it. The churches have an obvious role to play in this respect – if only they could see it.

The principle of reciprocity is popular because it unites giver and receiver in a mutually satisfying relationship in which the former can see the beneficial outcome of assistance, while the latter can derive a sense of achievement and self-confidence from meeting the standard of conduct expected. There is an infinite variety of ways in which reciprocity can be institutionalised. Here I would like to canvass just one particularly promising form of voluntary assistance, namely, microfinance, or the lending of small amounts of capital outside the formal commercial sector. This was pioneered as a means of development aid in 1976 by the Grameen Bank of Bangladesh. It consists of lending small amounts of capital to people who are too poor to interest the commercial banks, and relying on peer pressure from groups of loan recipients to ensure repayment. A woman might be lent as little as US$50 to buy, for example, a sewing machine and set up a microbusiness from which the loan is repaid. Amazingly, the repayment rate is higher than that for ordinary commercial loans. Eventually, many such people become wealthy enough to deal with the commercial banks, while the loans they have repaid with interest are lent to other people at the bottom of the income ladder.

One of the most active proponents of the microfinance movement is New Zealand-born David Bussau, who in 1979 set up Opportunity International after witnessing the failure of government-to-government aid programmes to help the poor after an earthquake in Bali. He sold his businesses and spent the next five years organising microenterprise programmes in Indonesia. Now, the Opportunity International network, based in Sydney, has over $30 million invested in over 100,000 jobs in 27 countries. It receives donations from the public, and governments are starting to channel their development aid through it and similar organisations in recognition of its success. It remains little known, but whenever it is publicised it attracts many inquiries from members of the public who are willing to make donations but want some assurance that the money will be well spent. The principle of using the money not as simple handouts but as loans that will be repaid and used over and over again is immensely appealing and needs no elaborate advertising to promote it.

Microfinance has also been pioneered in New Zealand as a form of domestic voluntary assistance. Kingdom Resources is a Christian non-profit organisation that helps low-income households in Christchurch and North Canterbury that are in financial difficulty. It offers them budgeting advice and if necessary reschedules their debts with interest-free loans; it also offers them a three-week course that aims to raise their members' self-confidence, trains them in managing their incomes, and helps them find employment. Members of the public can help in a variety of ways, not just by giving money but also by investing it in non-interest-bearing fixed-term or on-call savings accounts. Businesses can open accounts that take their monthly cash flows or set aside money for provisional tax. Kingdom Resources has 136 volunteer budget advisers and in 1996 advised a total of 592 clients.

These microfinance experiments show that assistance can be given in ways that break down the simple-minded categorical distinctions that so hamper our thinking about voluntary welfare. The initial distinction between givers and receivers becomes dissolved in a network of multilateral mutual aid. Altruism and self-interest are not in conflict but are both harnessed to increase the welfare of all. Reciprocal relationships contain elements of both exchanges and gifts. Comprehensive reform of the welfare state would create opportunities not just for the extension of assistance through microfinance but also for other imaginative and creative forms of charity yet to be devised and tested.

In helping the voluntary sector to do its job and reveal its comparative advantages over state welfare, we need not only to give more resources to it, but also, and more important, to pay more attention to the ways in which we give. As Robert Whelan tells us in the book I quoted earlier, most of us could devote more time, if we organised it properly, to directly volunteering our services and becoming active participants in the organisations that we support. The churches and the charities themselves could surely take the lead by spending less time lobbying the government for subsidies or welfare policy changes, and more time informing the public about the amounts and kinds of help that are needed. They could, for example, mount a campaign to persuade New Zealanders to emulate the American practice of giving away, on average, one week's income every year. Philanthrophy in America is not primarily associated with business, contrary to some views in this country: 95 percent of charitable donations in America come from private individuals and foundations. And rather than relying on the government's alleged comparative advantage in raising money, they could call for a return of the days when New Zealand was a truly more caring and charitable society, before the welfare state and high taxation took away much of the incentive and the means to make charitable donations.

In any case, it is far from clear that government has any comparative advantage in raising money for welfare. The administrative, compliance and other deadweight costs of taxation in New Zealand are at least 30c in the dollar: probably higher than the administrative costs of most charities, and certainly much harder to reduce. This is also an argument, incidentally, for helping the charities by means of general tax cuts rather than tax breaks for donations, since the latter imply higher general rates of tax and hence higher deadweight costs.

Just as we can no longer plausibly delegate our philanthropic duties to the state, so we can't pass the buck to 'the rich', whether directly or under the guise of the so-called 'social responsibilities' of corporations. Some of the best-known philanthropic foundations were created in the nineteenth century by people who were less wealthy absolutely but more wealthy relative to the average than anyone is today. As Peter Drucker has pointed out, J P Morgan's fortune at its greatest was about one third of Bill Gates's fortune today, in real terms. But whereas Morgan could finance all America's economic needs (other than residential housing) for four months, Bill Gates could finance America for only about two days. Just as the mass of middle income earners pay the lion's share of tax revenues, so they will have to finance the bulk of any shift from state welfare to private welfare. It really is up to all of us.

To conclude: the problem with much modern charity is not, as Brian Edwards claims, that it is offensive. To view the work of someone like Sister Connie Driscoll in this way is extraordinary. But charity does challenge socialism, a major motive of which has always been power-worship, or an identification with the state, and a corresponding resentment of the spontaneous actions of private individuals which may well undo the state's best-laid plans and reveal their inadequacies. Charity can be unreliable and capricious, as Edwards also said, but that is also true of state welfare, and probably more so. Certainly, charity can, like state welfare, thoughtlessly perpetuate dependency by merely relieving symptoms. Marvin Olasky tells in his book how he discovered this by pretending to be homeless. In March 1990 he dressed himself in the appropriate ragged attire and began to shuffle slowly around the streets. What happened?

If compassion is to be something other than a matter of self-congratulation, and charity more effective than immediate and impulsive redistribution, there is no alternative but to think harder about what we are doing. The American social historian Gertrude Himmelfarb has put it in a way that aptly sums up much that I have been saying: