We asked thousands of people to describe their ideal distribution of wealth, from top to bottom. The vast majority -- rich, poor, GOP and Democrat -- imagined a far more equal nation. Here's why it matters.
The inequality of wealth and income in the U.S. has become an increasingly prevalent issue in recent years. One reason for this is that the visibility of this inequality has been increasing gradually for a long time--as society has become less segregated, people can now see more clearly how much other people make and consume. Owing to urban life and the media, our proximity to one another has decreased, making the disparity all too obvious. In addition to this general trend, the financial crisis, with all of its fall out, shined a spotlight on the salaries of bankers and financial workers relative to that of most Americans. And on top of these, and most recently, the upcoming presidential election has raised questions of social justice and income disparities, bringing the issues into focus even more.
It is relatively easy to think about inequality as being too great or too little in abstract terms, but ask yourself how much you really know about wealth distribution in the U.S. For example, imagine that we took all Americans and sorted them by wealth along a line with the poorest on the left and continuing as wealth increases until on the right we have the richest. Now, imagine that we divide them into five buckets with an equal number of citizens in each. The first bucket contains the poorest 20% of the population, the next contains the second wealthiest tier, and so on down to the wealthiest 20% (see Figure 1).
With this in mind, from the total pie of wealth (100%) what percent do you think the bottom 40% (that is, the first two buckets together) of Americans possess? And what about the top 20%? If you guessed around 9% for the bottom and 59% for the top, you're pretty much in line with the average response we got when we asked this question of thousands of Americans.
The reality is quite different. Based on Wolff (2010), the bottom 40% of the population combined has only 0.3% of wealth while the top 20% possesses 84% (see Figure 2). These differences between levels of wealth in society comprise what's called the Gini coefficient, which is one way to quantify inequality.
This is the level of wealth inequality that exists in America, and it is clearly higher than people think, but in Goldilocks-esque fashion, we can ask: Is the real level of inequality too high, too low, or just right? When economists consider the desirable level of inequality, they usually define the ideal inequality from the perspective of economic efficiency: What level of inequality will motivate people to be the most productive and move up the wealth ladder? What level of inequality will allow those at the top to lift up society as a whole (say, by having the resources to invent new technologies)? What level of wealth will keep salaries low and competition high? And so on.
This is one approach to assessing the desirability of wealth inequality, but Mike Norton and I wanted to examine this question from a different perspective--that of regular (non-economist) people--and we wanted to examine inequality in terms of its effect on society as a whole, not just in terms of economic efficiency. After all, inequality is not just about economic efficiency. It's also about our day-to-day experience as citizens, the influence of envy, our social mobility, the importance of equal opportunity, our mutual dependency on each other, etc.
But what does it mean to ask people what level of inequality they want? And how do we get people to respond to such a question without being influenced by their current state of wealth? After all, wouldn't the rich want a higher degree of inequality and the poor want a more even distribution of wealth?
HOW MUCH INEQUALITY DO YOU WANT?
We took a step back and examined social inequality based on the definition that the philosopher John Rawls gave in his book A Theory of Justice. In Rawls' terms, a society is just if a person understands all the conditions within that society and is willing to enter it in a random place (in terms of socio-economic status, gender, race, and so on). In terms of wealth, that means that people know everything about the wealth distribution and are willing to enter that society anywhere along the spectrum. They could be among the poorest or the richest, or anywhere in between. Rawls called this idea the "veil of ignorance" because the decision of whether to enter a particular society is disconnected from the particular knowledge that the individual has about the level of wealth that he or she will have after making the decision.
With this definition in mind, we did two things. First, we asked 5,522 people to create a distribution of wealth among the five buckets such that they themselves would be willing to enter that society at a random place. Their answers could range from a perfectly even distribution with 20% of wealth in each quintile to a fully biased distribution with 100% of wealth in one and 0% in the rest.
We found that the ideal distribution described by this representative sample of Americans was dramatically more equal than exists anywhere in the world, with 32% of wealth belonging to the wealthiest quintile down to 11% by the poorest (see Figure 3).
What was particularly surprising about the results was that when we examined the ideal distributions for Republicans and Democrats, we found them to be quite similar (see Figure 4). When we examined the results by other variables, including income and gender, we again found no appreciable differences. It seems that Americans -- regardless of political affiliation, income, and gender -- want the kind of wealth distribution shown in Figure 3, which is very different from what we have and from what we think we have (see Figure 2).
We understood that setting up an ideal wealth distribution is a rather difficult proposition, so in another task, we made things simpler (see Figure 5) and asked people to choose between two unidentified distributions (again under the veil of ignorance). The first option, unbeknownst to participants, reflected the distribution of wealth in America. For the second option we modified the distribution found in Sweden, making it substantially more equal (we referred to this fictional nation as "Equalden").
We discovered that 92% of Americans preferred the distribution of "Equalden" to America's. And if one were to assume that the 8% who preferred America's distribution was made up of wealthy Republican men, he or she would be mistaken. The preference for "Equalden" was slightly different for Republicans and Democrats, and in the expected direction, but the magnitude was very small: 93.5% of Democrats and 90.2% of Republicans preferred the more equal distribution. While this 3.3% difference is substantial when we think about the economy of an entire country, if we look at it from the perspective of the gap between Equalden and the U.S., it's clear that the similarity across the political spectrum is far more substantial than the differences. And once again, participant's gender and income level did not produce any appreciable difference in this preference.
LEARNING FROM (THE VEIL OF) IGNORANCE
There are a few lessons that we can learn from this. The first is that we vastly underestimate the level of inequality that we have in America. Our society is far more uneven in terms of wealth than we believe it is. Second, we want much more equality than both what we have and what we think we have. Apparently, when asked in a way that avoids hot-button terms, misconceptions, and the level of wealth people currently possess, Americans are actually in agreement about wanting a more equal distribution of wealth. In fact, the vast majority of Americans prefer a distribution of wealth more equal than what exists in Sweden, which is often placed rhetorically at the extreme far left in terms of political ideology--embraced by liberals as an ideal society and disparaged by conservatives as an overreaching socialist nanny state.
A third lesson concerns the political gap between Democrats and Republicans: Given the extraordinarily polarized and derisive rhetoric flying back and forth between Democrats and Republicans, one would think there was an insurmountable gap between their positions. So how is it possible that we found so little difference between them in our study? One reason for this could be our inability to separate our ideology from our current state of wealth. Our interests tend to color our view both of how things are and how they should be. Another reason could be politicians, who, in order to rally people to their side, try to generate feelings of greater difference and opposition--and therefore conflict--than actually exist. From this perspective one could claim that politicians obfuscate similarities by using galvanizing but elusive terms like "small government," "tax relief," and "freedom."
Rawls' veil of ignorance deals with such superficial and irrelevant influences on what we think by prompting people to consider all possible socio-economic situations rather than just their own and the interests and ideologies that come along with that. The veil of ignorance accomplishes something similar to blind taste testing. Take wine, for instance. If a person knows the appellation and price, and realizes that French wine is usually preferable to Finnish, his or her perception and opinion of how good each wine tastes will be influenced by these preconceived notions. Similarly, when we express opinions about politics and life in general, we can't help but be influenced by our own varying degrees wealth and ignorance of others' lives. The veil of ignorance works to separate our core beliefs from the biases and prejudices we develop over time and through the subjective experience of being part of a certain class and demographic.
As for what this means about changing the level of inequality, which from our study seems almost unanimously objectionable, there are essentially two paths: education and taxation. Improving education works in a sense to change the input into the economy--better-educated workers are more resourceful and employable, and can move up the economic ladder. Changing taxation deals with the output--those who prosper pay more into the system than those without the same benefits. Our study doesn't tell us anything about which of these two approaches to reducing inequality would be preferable, but in practical terms, bridging the huge gap between what we currently have and what we want to have would require a mixture of both. Our study also doesn't deal with how to bring what people say they want under the veil of ignorance into line with what they're willing to do when it's their money and resources that are about to be distributed. It is one thing to get people to tell us what kind of society the would want to join, and another to get them part with their money in order to create that society.
With all these objections in mind, it still seems that the political discourse could benefit from a Rawlsian approach. Taking this path could help us understand what it is that we really want and then allow us to consider ways to get there. If our politicians also accepted this starting point, they might argue less about ideology and differences and more about paths to get closer to our common goal--a much less extreme level of inequality.
Social justice and optimal wealth distribution are highly complex topics, and it's hard to imagine that any study could dramatically change opinions about education, welfare, or tax reform. But consider this. When we ran the same basic experiment in Australia, we found Australians did not differ much from Americans in their views of the ideal distribution. When we ran another version of it with NPR listeners, and then readers of Forbes Magazine, the results were still basically the same. And most likely, if you participated in one of our tests, your response too would have fallen in line with these findings.
So whatever you think the current state of wealth distribution is, and whatever you believe the ideal level of wealth distribution to be under the veil of ignorance, there probably is a gap, and a large one, between the two. Awareness of the disparity between what we have and what we want implies that, ultimately, we as a society need to face the problem and find a solution.
Wolff, E. N. (2010). "Recent Trends in Household Wealth in the United States:Rising Debt and the Middle-Class Squeeze--an Update to 2007." Levy Economics Institute of Bard College.
Norton M.I. and Ariely D. (2011) "Building a Better America--One Wealth Quintile at a Time." Perspectives on Psychological Science., 6 (9), 9-12.
Rawls, John. (1999). A Theory of Justice. Cambridge, Mass.: Belknap Press of Harvard University Press.