There is, to the great grief of the old-style interventionist economics, no such thing "out there" as total utility since the "utilities" of different persons do not add up to a homogenous quantity. The total of a cherry and a prune is a cherry and a prune and never two cherries nor two prunes. At best, one can make the value judgment that it would be better to have two cherries and no prune. However, the phoney arithmetic of interpersonal aggregation combined with the misunderstanding about the "diminishing marginal utility of income" (which may or may not work for a person but does not work across persons) creates the belief that redistribution is not only a moral imperative, but also makes a society "objectively" richer.
In tirelessly battering "neo-liberalism" (but why "neo"?) such eminent economists as Joseph Stiglitz and their hordes of followers condemn "globalization" as a scheme for making the rich richer and leaving the poor by the wayside if not actually making them more miserable. On the part of the eminent economists, such claims show how passion and indignation can cloud logic and the perception of evidence. On the part of the horde of camp-followers, they show for the umpteenth time that a little economics is worse than none.
Two fallacies support the demand that incomes on a world scale ought to be made more equal. One is that just as wealth creation is good per se, inequality is bad per se. Since "globalisation" jointly produces wealth and inequality, a good and a bad, a balance must be struck between them. Maybe "globalisation" is bad on balance unless its inequality-generating effect is "corrected".
The other fallacy is rooted on the recognition that since the age-old process of "globalisation" has gathered such momentum in the last two decades or so, factor shares in developed countries have shifted in favour of capital. Except for the highly skilled, labour incomes fell behind economic growth. Seeing the relative share of capital increase, and the "multinationals" piling up the billions, the cry went up that the rich are getting richer. However, the day when capital was mainly owned wholly or even mostly by the rich is long past. Today, the "multinationals" are almost without exception majority-owned by middle and low-income people through pension funds and other intermediaries. The benefit accruing to them does not show up in the statistics of personal incomes, or does so only after a long time lag. Changes in the share of profits vs. the share of wages do no longer clearly signal a change in income distribution between rich and poor.
Where inequality has become glaringly greater and more obvious is in the remuneration of top corporate executives who deploy other people's assets in the markets for goods and services, and in that of fund managers who deploy them in financial markets (though the latter is probably no more than a passing aberration that should soon correct itself, while the former depends on whether shareholder activism will gather much needed strength). These somewhat perverse results of the principal-agent conflict are plainly offensive and in bad taste, but in terms of total national or world income not as big as the noise they provoke.
As regards the large-scale problem of income equalisation, the theoretical perspective could hardly be simpler. When the mean income exceeds the median, a majority of a self-interested electorate should vote for rich-to-poor redistribution, and go on doing so until income distribution attains equality. (At this point, any further redistribution would start creating inequalities, though that is no reason for expecting it to stop).
Any government that would falter in the progress toward complete equality should be voted out and replaced by its rival that promised to go further. How come, then, that there is no trace of any strong trend toward income equality? To all appearances, governments and their oppositions fight their main battles over redistributive measures, vast programmes to that effect are constantly born or reshaped, yet the statistics do not show that inequality is being ironed out.
In abstract theory, the competition for capturing a majority of votes takes the form of groups of candidates for office making bids to pay voters with below-the-mean incomes sums of money taken from voters with above-the-mean ones, the highest bidder(s) forming the government. Redistribution is a neat rich-to-poor money transfer. In reality, however, the scope for such transfers is very limited, squeezed up as it is by three huge blocks of in-kind transfers governments make from rich and poor to rich and poor, with little net redistributive effect. They are old-age security (pensions), health care, and education. They are financed by general taxation and compulsory social insurance (payroll taxes). Both rich and poor contribute to them, income taxes weighing more heavily on the rich, value added or sales taxes and petrol taxes more heavily on the poor, with the incidence of payroll taxes being hard to assess but likely to hit the poor harder. (Corporation taxes should be regarded as part of personal income taxes paid by corporations on behalf of their ultimate owners).
In advanced Western-style economies, pensions will absorb 15-18 per cent of GDP, health care 7-12 per cent and education 5-9 per cent, a total of 27-39 per cent. Not all of this will be financed by taxes, compulsory social insurance premiums, and deficits (borrowing), but most of it will be. With public administration, defence, law and order, and the service charge of the national debt also to be financed, the scope for straightforward rich-to-poor transfers, as envisaged in pure theory, is obviously constricted. Willy-nilly, the state takes on in-kind, paternalistic commitments in old age and health care and education that bind its hands but that it cannot significantly lighten without hopelessly spoiling its image as the caring and civilising guardian of the common good.
Redistribution of the sort that reduces the measure of income inequality and strongly favours the poor is popularly called "social justice". Naming it so is a clever and audacious stroke of one-upmanship, for the words, though strictly meaningless, pre-empt for it the moral high ground that rightly belongs to justice. However, the redistribution that goes on in modern economies is of a different sort. It leaves inequality mostly unaffected, though it absorbs up to (and in some cases more than) one half of people's incomes in order hand it out again in various in-kind benefits to much the same people. It is, in plain language, "churning" the national income without there being much evidence that all its mixing, shaking and heaving is leaving any identifiable class of society noticeably better off.