Hirsch first discussed the importance of positional goods, a term he coined, in the Social Limits to Growth (1976). A review of that book by Jack Bakunin is available at "The Failure of Individualism," maintained by Religion Online.
First, consider what are known as positional goods. As first elaborated by the economist Fred Hirsch, positional goods are those products and services which are inherently impossible to mass produce because their value is mostly, if not exclusively, a function of their relative desirability. Consider a simple ranking of the best restaurants. Assume for simplicity that the best restaurants are the most fashionable and most desirable eateries. However good the general run of restaurants are, the most favored top shops are still better, more desirable, and more exclusive than the others. Imagine that all restaurants were to improve by proportionate degrees, hence raising the average quality still further. But no improvement of the average will change the fact that only ten restaurants can—by definition—make it to the list of the ten best restaurants.
Now consider what happens as incomes and population grow. If the reason to go to the best restaurants is to be seen at the "best" establishments it will become no easier and may get harder to get into the top ten. Assuming for a second that only the wealthiest get spaces at the top restaurants, no increase in average income will change the fact that only the richest will be capable of going to those establishments. The ordinary joe may get wealthier, truffles may become as cheap as corn, and a fast food joint may come to have food that would dazzle today's gourmet if technology and productivity continue to promote improved food. But no increase in average wealth will make the best any more available to the masses.
For those who only care about getting a good meal, this is a blessing not a tragedy. But those with expectations of going to the great restaurants as their incomes rise will be frustrated by the fact that the best remain agonizingly out of reach unless they grow rich much faster than the average. So only the wealthiest of the wealthy can have these goods and must pay a growing premium to do so.
While some goods depend on the value of having the "best," another type of positional good is one where the value is due to its irreproducibility. This is especially true for fashionable goods that are literally fixed in numbers—say the paintings of Picasso. The price that will be paid for these goods will depend on demand. And demand is partly a function of the size of the population wanting a Picasso and the relative wealth of that group. If you are at the bottom of that group now, you will be no more able to obtain the Picassos even if your incomes rise and even if the absolute differences in income between yourself and the top demanders grow increasingly smaller. So long as the top few bidders can buy the hundred most wanted paintings, they will continue to be those who get those paintings, whether the gap between the hundredth and the hundred and first bidders is large or small.
But what this means is that the relative price of these non-reproducible items will be driven up as the price of other goods falls. Consider the one positional good that is of greatest concern to most groups of people—housing. It is a commonplace to hear people complain that housing has become more expensive and less available. But the rise in housing prices stems from three rather different phenomena. The first and most obvious is inflation. The nominal price of housing may rise even though its real price may have fallen, in much the same way that gas at $1.25 a gallon today would be cheaper than gas at $.25 a gallon fifty years ago. The second point is that in comparing the average house of today with the average house of twenty, forty or a hundred years ago, we are mixing apples and oranges. For one thing, the average U.S. home today is much larger than in previous decades. Moreover it is more likely to have heat and even air-conditioning, central plumbing, storm windows, and a garage than "equivalent" houses of previous times. Hence, the average U.S. house today, is better than the average house of yesteryear.
Housingnyc.com provides an idea of current prices for apartments around New York City, including Manhattan. The figures shown are median prices, not the high end. To get a feel for how expensive the high-end apartments are, see the "Manhattan's Grandest Apartments" section of this report from Brown Harris Stevens, a luxury real estate company.
But over and above these is the third problem: location. And good location is the quintessential positional good because it is important both for its relative prestige and its irreproducibility. When the upper middle classes complain about housing prices, they are really fretting about the cost of housing in the most desirable locations. But this exactly conforms to the phenomenon we've been discussing. If the ten best locations in, for example, Manhattan, always go to the ten wealthiest residents, then no others can have those locations no matter what the average incomes are. Rising incomes and population growth will disproportionately drive up the price of such homes, frustrating those with dreams of upward mobility but ultimately causing the wealthiest to pay more for locations that they would have gotten anyway, had everyone else's incomes not grown so much. Correcting for quality and inflation will show that the best houses in the most desirable locations are expensive primarily because of increased demand for housing in the best locations. This is not just because of the prestige of good locations. Good locations have "objective" benefits (such as access to the centers of commerce and finance) that become more valuable with population growth and economic progress even if one is indifferent to the "status" value of good location. The average U.S. house might be located in communities that are better (due to economic and social improvements) than most of the prosperous neighborhoods of a century ago but the number of top locations will remain unchanged. Indeed the best locations will become disproportionately more expensive. Thus, absolute wealth will not necessarily buy you the best places, while the wealthiest of the wealthy will end up paying more for locations they would have obtained at lower prices if people were less prosperous in general.
Though Hirsch decried the "waste" that competition for positional goods like location creates, I see it as a positive force for the democratization of the benefits of economic growth. Thus the spending by the wealthy on many positional goods acts as a curious sort of natural taxation. The richest (or most ambitious) must work harder and pay more for virtually the same goods as yesteryear while their productive investments (necessary to stay on top of the income distribution) benefit the entire economy.
The second class of goods whose prices rise disproportionately are those dependent on human labor. The flip side of high average incomes is expensive labor costs. Hence the goods that directly or indirectly rely on human input become more expensive as growth progresses. In many countries around the world, even the middle classes can afford to have live-in maids. As incomes rise and the opportunity cost of performing menial labor changes, fewer people will want to work as servants. Those that do will command a higher price. So in America the middle and upper middle-classes are less likely to have servants than in much poorer nations.
Higher labor costs due to rising incomes also affect the supply of goods requiring specialized knowledge and high degrees of craftsmanship. Those members of the population with a taste for hand-carved furniture, stained glass windows, or other items requiring patience and craftsmanship will find that the number of people willing to work at such tasks may shrink and the prices that are charged by those that remain may grow. In certain cases, the costs may so discourage demand, and the supply of workers may fall below a critical level so that even the specialized knowledge itself is lost to future generations. Again, this change in relative prices is likely to be felt most acutely by the wealthiest, but it will also affect those of more modest means who have a peculiar affinity for some items that they feel are essential to the good life.
Now, all these issues have not gone unrecognized in the literature, but in my view, insufficient attention has been paid to how they cloud what might otherwise be reasonable debates about current and future policy. Unlike most discussions of not so easily reproduced goods, this essay argues that their continued presence guarantees the persistence of inequality no matter how much economic progress is made or how goods are distributed. In fact, one might say that successful economic growth will inevitably lead to the perception of important and mostly ineradicable inequalities. In a world where all goods are free but one, differential access to that last good will loom large in public consciousness regardless of what the nominal measures of dollar or material inequality tell us.
Whether one sees this as a good or bad thing depends on any number of factors of course, not least of which is one's attitude towards the desirability of promoting or ignoring material equality. If your goal is to narrow the gap on most important material dimensions, that world will be a good one, no matter what the numbers show. But if you view the perception of inequality as an important subject for policy intervention, then the future is grim. Not only will redistribution of the fruits of progress not erase such gaps, they will foster more dissatisfaction with inequalities that remain. The greater the success in the economic sphere, the greater the conflict over inequality. Indeed, the most serious problems will arise from goods which cannot be purchased in the market at any price.
It is worth recalling that on nominal measures, the old Soviet Union did not have especially high income inequality. But the important differences were those that money could not buy. Political connections or talents blessed by the state (e.g. ballet or chess) were more important determinants of what you could buy or where you could shop than simple monetary income.
But of course, in a future world where accidents of birth and the fortunes of good genes are even more critical determinants of success than they are now, inequalities that persist will be especially galling. Even if biology could somehow be conquered to the point that genetic good fortune could be parceled out equally to all, the minor differences that remained would loom ever larger. Assume that the only thing that differentiated people were beauty (ignoring the role of cosmetic surgery and better make-up). Then those blessed with desirable features may come to seem vastly more privileged than the rest.
Moreover some people have always measured their success in life by their ability to control and command other human beings—whether as servants, serfs, or soldiers. Perhaps this power gap may come to be the one source of unequal income that really matters. Whether you view such an eventuality as desirable or irrelevant, more of our intellectual effort should be devoted to this future scenario. Not simply because we are heading there, but in many ways because parts of that world are already here. Because social and positional inequalities already distort existing measures of incomes and wealth, many seemingly clean cut economic debates are more intractable than one would imagine. And of course, social anxieties over the unavoidable differences will become even more troubling, the less we can constructively address these issues. By ignoring the role of the goods that are inherently limited or irreproducible in our measures of simple material inequality we inevitably end up discussing the wrong issues. It is likely that as differences on important margins narrow, the most unequal efforts and expenditures will accrue to those margins that remain difficult to change.
A brief discussion of income inequality trends in different parts of the globe can be found at the World Bank website.
Even today, we routinely exaggerate the extent of material inequality and make foolish comparisons between different time periods and between countries at different levels of development. This does not mean that inequality has disappeared, or that it is unimportant. But in many cases (such as the plight of children in the poorest countries) the real problem is not unequal distribution, but lack of the most basic goods that have long become commonplace in the developed world. Of course, even with the reminder of other people's misery, the small differences that seem close to us often appear large enough to block our vision of the bigger picture. I wish it were otherwise, but I foresee both increased prosperity and still louder complaints in our future.