The topic du jour on the econ blogs is Krugman's claim about the effect of politics on income inequality. Before I solve that problem for you in different mindblowing post, let me say that I don't really understand why economists care about income inequality qua economists. I understand why they care about income: more money buys a more preferred consumption bundle, which, by definition (not mechanism) gives you more utility.
What leaps to mind is that economists must be impressed with diminishing marginal utility (DMU) arguments for redistribution. But this is not qua economist. Orthodox utility theory is not in the business of interpersonal utility comparison, so there is no language in which to say that overall utility has been increased by redistribution. There's just one guy's utility going down, another guy's going down, and no way to say whether one compensates for the other. (This is one of the reason some economists are attracted to happiness research: they want a language in which to do principled interpersonal utility comparisons.) But, as I have noted in the past, when economists have anything interesting to say, it's usually because they are applying what I like to call economic folk morality, which plays fast and loose with the various senses of utility. Economists qua amateur moral philosophers in my experience are very impressed with DMU justifications for redistribution.
But professional moral philosophers with high-level training in economics, such as David Schmidtz, are rather better at this kind of thing than economists who occasionally indulge their philosophical prejudices on blogs and in the NYT. Schmidtz shows that even assuming everyone has the same utility function, marginal utility diminishes smoothly, there are no incentive problems, and redistribution costs nothing, it still does not follow that utility must increase when inequality is reduced.
Schmidtz notes that DMU is precisely why some dollars are not consumed, but invested. There is a point where an extra dollar will have a higher expected utility if allocated to production than consumption, and we need enough people at that point to have the kind of production that ensures that there is anything to redistribute. From Schmidtz's Elements of Justice (which I highly recommend), where he discusses a simple example involving the redistribution of corn:
Precisely because of diminishing (that is, downward-sloping) marginal utility of consumption, production becomes a higher-valued use as wealth … rises.
Note: Production's tendency to be more desirable relative to consumption is a general consequence of consumption's DMU, and not an artifact of an odd example. The general conclusion: If a community does not have people out that far on their utility curves, so that they have nothing better to do with marginal units of corn than to plant them, the community faces economic stagnation at best.
Therefore, unequivocal utilitarian support for egalitarian redistribution is not to be found in the idea that consumption has DMU. This result in no way depends on questioning the premises of the DMU argument. On the contrary, it is grounded in DMU. A society that takes Joe Rich's second unit [of corn] and gives it to Jane Poor is taking that unit away from somebody who, by his own lights, has nothing better to do than plant it and giving it to someone who, by her lights, does have something better to do with it. That sounds good, but in the process, the society takes seed corn out of production and diverts it to current consumption, thereby cannibalizing itself.
There is much more to Schmidtz's argument, and I encourage you to check it out. Like Schmidtz and like a bad socialist, I am cold on material equality, and, like a good liberal, warm on material sufficiency. What matter is that people have enough to realize the ends that make our lives worth having. If a system that encourages those who already have enough to invest their extra in production is more likely than a heavily redistributive system to provide everyone with enough, then a highly materially unequal society may best serve liberal ideals.
Whether we should be morally worried about inequality depends on its sources. But then the real worry is the source of inequality, not inequality per se (though the resulting inequality may exacerbate or consolidate the original injustice.) If inequality is based in predation, then the moral worry is predation. In a system like ours rife with corporate welfare, some people do get rich off political predation. (I can see K St. from where I'm sitting!) I don't know how many of the top 1% (that both includes Krugman and makes him insane with indignation) got there through rent-seeking. Maybe a lot, but certainly not all the software millionaires, and all those entertainment and sports stars. Anyway, as I've mentioned before, redoubling our efforts Krugman-style to take more money from the rich through the political system is mostly an effective way of providing the rich with an incentive to put more money into the political system to prevent having even more taken. It is in principle possible for the people with the least leverage and weakest incentive to bend the political system to their will, just as it is in principle possible for a dog to catch its tail. The only option for separating money and political power is to reduce the desirability of political power by reducing its scope and effectiveness. My guess is that constraints on government power, such that the rich would have less incentive to try and dominate it, would lead to more people having enough, but perhaps even greater inequality.