Reading Time: 4 minutes

Conover et. al. presented at the 2011 ICWSM conference

Image from ‘When Computation met Society’

My posts regarding the Schwartz-Duval Model of Human Values focus on the values that people hold, and how those values relate to their politics. While it is important to understand the differences between the values that underpin people’s politics, it is equally important (some might say more important) to recognise the similarities. So today I’m going to focus on a paper from 2011, by Michael Norton and Dan Ariely, that looks at preferences in income inequality between various demographic sub-groups in the US.

Building a Better America—One Wealth Quintile at a Time

There is a presumption amongst many liberals that conservatives want to increase inequality. The perennial calls for tax cuts for the rich, being the go-to example. Equally, there is a presumption amongst many conservatives that liberals want equal outcomes in income (and everything else) for everyone (cries of socialism, Communism, and Marxism, etc.). These are of course incorrect for the vast majority of those who self-identify as either one of the political types. Indeed, the vast majority does seem to recognise that there is inequality and do want to see the balance redressed somewhat.

Norton & Ariel figure 1

As we can see, the vast majority does indeed want a Sweden-like balance to income. Part of the reason why our views of each other are so distorted is that stereotypes rely on the difference. This similarity, while known by the individual, doesn’t form part of the individual’s stereotypes. So while there are differences between liberals and conservatives (and other demographics), the difference is not the chasm that it may sometimes seem.

But, if our view of each other’s perspectives is stark, our view of the reality of the situation is not stark enough. (Remember that this paper is the best part of a decade out of date, and the trend is for the inequality to worsen.)

Norton & Ariel figure 2

So what of the demographic split? Females prefer the Sweden distribution over the US distribution (92.7%), but males do, too (90.6%). Bush supporters in the 2004 US presidential election (90.2%) were outnumbered only slightly by those who voted Kerry (93.5%). In the case of income, the demographic most directly affected by whether or not such preferences were somehow enacted, are also not as stark as one might expect. Those earning less than $50,000 preferred the Sweden Distribution (92.1%), as did those earning between $50,001 and $100,000 (91.7%), and those earning more than $100,000 (89.1%).

Norton & Ariel figure 3

The vast majority of people are in approximate agreement about the direction of required change, and apparently unaware of the magnitude of the problem. It seems likely that beliefs about income (in)equality are rooted in our beliefs about fairness (and probably harm, too). As we know from Jonathan Haidt’s work, liberals tend to value redressing harm and unfairness more, as such liberals will likely notice and be more upset by this difference. Conservatives do still value these things, just to a lesser degree, being concerned about issues of In-Group, Authority, and Purity/Sanctity, too.

If we mostly agree, what’s the problem?

According to Gilens & Page (2014), the lack of progress in redressing income inequality (and all other political ends) has nothing to do with popular preference (see below).

Gilens & Page (2014)

Figure 4 (from Gilens & Page, 2014): Predicted probability of policy adoption (dark lines, left axes) by policy disposition; the distribution of preferences (gray columns, right axes).

It also seems clear that a sizable chunk of the beliefs about the other political persuasion to one’s self is rooted in the actions of the politicians that “represent” them, rather than the individuals in question (but such is the way with stereotypes).

Finally, pointed questions

Is my experience unique? I hear a lot more complaints from liberals about liberal politicians than I hear from conservatives about conservative politicians. I also encounter a lot less recognition of this, coming from the right, than I do the left.

To give an example; whenever I used to venture onto The Young Turks, on YouTube, there were a lot of complaints by self-professed libertarians and conservatives about how partisan the show was. However, at the time, every other piece that Cenk Uygur did was a complaint about some shortcoming of Obama (e.g. the gap between his campaign promises and his enacted policy, his watering down of more progressive policies in an attempt to reach across the aisle, etc., etc.). By comparison, it was sufficiently rare to hear Fox News say something negative about a Republican, that it often made the news itself, or at least made waves in the liberal blogosphere and social media.

If the problem is politicians and how easily money and apparent influence (as proxies for popularity, if we are being generous to them) sway them, why do we continue to assume that they speak for their constituents?


To clarify my pointed question in response to Fragarach’s point in the comments:

For a liberal, it is no problem to criticise the person they support, but conservatives seem to have a harder time with this. For example, I could say that I admire Obama, but was unhappy with the drone strikes, TPP, Guantanamo, etc. And by the same token, I can say that I loathe Trump, but appreciate the nixing of TPP. My experience with many rank and file conservatives, and some Tea Party members, is that they are 100% for X, until X does something they can’t stomach (like losing), and then suddenly the are 100% against X, but 100% for Y (X and Y still being conservative politicians, but maybe from slightly different factions).