More articles that you could read in your lifetime demonstrate that socioeconomic status correlates with important outcomes, from voting to longevity. But what is SES? The American Psychological Association’s Task Force on Socioeconomic Status says (PDF):

social status is commonly conceptualized in terms of socioeconomic standing derived from formulas, taking into account various combinations of income, education, and occupation … Although social scientists continue to disagree about how best to operationalize SES, which indicators are the most valid (e.g., occupation vs. education vs. neighborhood), and the translation of different combinations of these indicators into class groupings (e.g., college degree plus corporate position equals “middle class”), the fundamental conceptualization

involves access to resources.

The Task Force proceeds to complicate the picture by introducing conceptions that are less focused on absolute resources and instead emphasize relative standing or the ability to reproduce advantage from one generation to another. But all the conceptions seem to presume that the SES of an individual can be captured by a single composite measure which is not equivalent to wealth, income, or occupation and which does not incorporate race, gender, or age. (Those are treated as correlates of SES, not components of it.) To know whether the formula is valid, we must decide what one thing we want to measure.

Perhaps SES is a measure of the comparative advantage that A has in getting what A wants in life, once we account for personal aptitudes and character traits, demonstrable skills, and race, gender, and age (which we measure separately).

We know, for example, that US Senators have median assets of $1.7 million and that 51 members of the last Congress had family members who had also served in Congress. So it seems likely that SES affects (but does not completely determine) your odds of getting into Congress–which is something that some people want.

But what we want is affected by SES, not just whether we can get what we want. Perhaps, for example, being related to a Member of Congress not only makes it easier to be elected; it also makes it more likely that you want to be a politician. (Certainly, many Americans would rather be almost anything else.) To take another example: you need wealth and family connections to be admitted to certain snobby clubs. But having wealth and family connections might hurt your chances of hanging out with the cool kids behind the gym or going ice-fishing with the guys. In short, SES (as we typically measure it) confers advantages on those who want to get into high-SES social circles and makes them more likely to want such entree. But that doesn’t mean that SES increases the odds of any person, A, to get whatever A wants.

Now it is starting to seem as if SES is not a continuous variable at all. “High SES” is just the name we give to certain subcultures, although other subcultures are equally exclusive and desirable. But that can’t be right, because it’s obvious that getting into the US Senate or the World Economic Forum at Davos is more valuable than being included with the cool kids behind the gym (even if most people would rather be with the cool kids).

Another possibility: SES is the most refined measure of economic power. The idea is that some people can influence their own and others’ economic circumstances more than other people can. (They can also use their economic power for political leverage.) Wealth is one measure of how much economic power individuals have. If you have a billion dollars, you can buy a company and fire everyone: that is power. But wealth is not the best measure of economic power, as a couple of examples will demonstrate:

• It is 1974 and George W. Bush is a graduate student at the Harvard Business School. He is an alumnus of Yale and a member of Skull & Bones. His father is chairman of the Republican National Committee and is on the road to becoming president of the United States; his grandfather had been a United States Senator and a board member of Yale, CBS, the Union Banking Corporation, and other companies. George W. Bush has high SES, if anyone does. But his income may be quite low because he is a graduate student, and even his wealth may be negligible if the Bush family’s assets are structured in such a way that he has to wait for his share. An auto mechanic of the same age may have more wealth.
• A 95-year-old widow lives in a house worth $2 million, but she can only use that $2 million asset as a place to reside. She has less power than George W. Bush had in 1974, but more wealth.

Yet another possible definition emerges: SES is an estimate of how much money the individual will be able to invest and spend according to choice during the rest of his or her lifetime. That definition appropriately gives a current Harvard MBA student higher SES than an auto mechanic or an elderly lady in a big house. But it won’t quite work, because not all “socioeconomic” advantages are monetary. For instance, having a grandfather in the US Senate is an advantage that money cannot buy.

If we want to measure economic power rather than wealth as SES, then such factors as age, educational status, parental occupation, family reputation, and family income should also count as aspects of SES.

Those factors could be included. But geographical location, language, race, gender, physical appearance, and religious belief–not to mention personal traits and accomplishments–also confer economic power. I understand the value of measuring these factors separately from SES so that we can investigate the changing relationships between race or age and SES. But if SES is defined as economic power, and if being white confers economic power in the US, then on what conceptual basis can we exclude race from our measure of SES?