Content Editor: Adam Daroff

Wednesday, March 15, 2017

What happens when America gets concerned about inequality?

An article of mine on the political implications of American concerns about economic inequality was just published in the scientific journal Political Behavior. The actual article is really long and technical, with lots of math and academic jargon, but here it is in a nutshell.1

What happens when Americans get concerned about economic inequality?

Economic inequality has been getting worse for the last few decades. You can see how the percentage of total income earned by the top 1% has been increasing since at least the 1980s, and the Occupy protests have apparently done nothing to slow it down.

A number of researchers have been trying to figure out why we as a country don't seem to be doing much to respond to increasing inequality.

For many of us, there seems to be a pretty obvious explanation: The "1%" have been using their power and influence to keep the government from stepping in and doing something about inequality.

This is precisely what many social science researchers think is going on. These researchers argue that many Americans are really concerned about inequality, and want the government to step in and do something about it, but the government isn't listening. These researches think that influence of the rich and powerful is preventing government policy from reflecting the “will of the people.”

A key part of this argument is the claim that, when Americans are concerned about inequality, that means they want the government to do something about it. Some researchers who have tried to test this assertion even think they've confirmed it.

But my research shows that they're probably wrong. Americans can get more concerned about inequality without getting any more supportive of government action. In fact, they might actually become less supportive!

Why do I think previous research is wrong?

Two problems.

1) Most Americans don't really spend a lot of time thinking about political issues like inequality, and many don’t have a real opinion on it one way or the other. So if you just go up to an average American and ask them “are you concerned about economic inequality?” and they say “yes”... that doesn’t actually mean that they are concerned about inequality! They might have answered “yes”  because were just laid off from their job, or “no” because they just saw a funny video on YouTube. We know from experiments that this happens a lot, especially when the question is about something complicated and abstract like “inequality.”

Political scientists have learned that, if you want know what Americans think about complicated, hard-to-define political issues like “economic inequality,” you need to look at a bunch of different questions that all ask about the same basic thing (like "inequality") in slightly different ways. By looking at these questions as a group you can separate the “signal” of real opinions out from the “noise” of people just saying “yes” or “no” for random reasons. Past work studying American views on inequality had not done this, but I did.

2) Other researchers claimed that being concerned about inequality caused Americans to become more supportive of government action, because they noticed that the same kind of people who were concerned about inequality also tended to support government action. But some Americans might just want the government to do more for some other reason and, because of that, they then decide to become more concerned about inequality. That might seem illogical, but Americans can be illogical sometime.

Maybe you’ve heard the statistical rule-of-thumb: “correlation is not causation.” Well, that’s not the problem we have here. The researchers were able to show that the relationship they saw really was (probably) some kind of causation, and not just correlation. The problem is that they couldn’t prove that the causation ran in the direction they thought it did.

The way to get out of this jam is to look at how people’s views change over time. If I see someone first become more concerned about inequality, and then becoming more supportive of the government doing stuff, I can be sure that it increased concern that caused the increase in support, and not the other way around.

What did I do?

To make a long story short, I made this chart, which shows how concerned Americans were about inequality, every year from 1966 to 2015:
Because of the way I combined different questions to create a stable measure of American concerns, the specific values on the Y-axis don't mean very much, but all you need to know is that higher values represent periods where more Americans were concerned about inequality, and lower values are periods where they were less concerned. The important thing here is not the level of concern, but how it changes over time:

  • Concern rises during the 1960s and 70s, during LBJ's "Great Society" and the "War on Poverty"
  • It takes a nosedive right before Reagan takes office (perhaps signaling the "Reagan revolution")
  • It immediately starts climbing again during the 1980s, as people begin to get fed up with "Reaganomics"
  • Concern declines during the 1990s - a period where broad-based economic growth benefited a large portion of Americans.
  • It's pretty flat during the the early 2000s – maybe people were more concerned about the war and terrorism than inequality. 
  • Then it starts to move up and down really fast.
My own view is that fast movement is probably just noise - because the data that I used to create this chart is kind of spotty after 2010. But there is an interesting spike that happens in 2011 - right as the Occupy movement shows up - which only lasts a year before concern declines again.

When America becomes more concerned about inequality, does it then become more supportive of the government doing stuff?

Past research said yes. But to answer this question myself I compared my own measure of concern with another measure (created by political scientist James Stimson) of how much Americans support government intervention - which political scientists call “policy mood.” Here’s what we find:

Once again, ignore the numbers on the Y axis and look at how the two measures change over time.

For example, notice that in the late 1980s “concern” and “policy mood” are moving in the same direction, but in 1970s they move in exactly opposite directions.

Using statistical modeling, I explored these patterns more carefully and found that there was basically no relationship between concern and mood. When Americans become more concerned about inequality, there’s no reason to suspect that they’re going to get any more supportive of the government doing stuff.

In fact, when I look more carefully, I find that there might actually be a tendency for Americans to become less supportive of government interventions after they become more concerned about inequality.

What does this mean for activism?

My research suggests that just going around telling everyone “inequality is a problem!” may not, in itself, do much to actually help reduce inequality. Even if we’re successful in getting people to become concerned about inequality, they might just decide that the best way to do something about inequality would be to reduce taxes on the rich, and get rid of government “red tape.” That might not make much sense to us, but that’s how some people apparently think.

This may have been exactly the problem with Occupy. My view is that by just shouting “inequality is bad” and shutting down any serious discussion of what to do about it Occupy blew its shot to actually get anything positive done, and may have produced some unintended consequences.

This research was the focus of my dissertation, which I submitted back in August of 2016, well before the election. Near the end I wrote a sentence that now seems a little eerie:

“Indeed it seems possible that “giving the people what they want” with respect to inequality could lead to a President like Donald Trump, who is able to rhetorically connect concerns over economic uncertainty and inequality to an entirely different set of policy positions, or else to a charismatic or symbolic rhetorical style devoid of any substantive policy recommendations.2

It’s a little late to stop that from happening. But maybe we can use these results to keep from making the same mistake a second time.


1Those of you who are also in academia can read the full article here....don't get me started on journal publication practices. While I’m at it I should mention that pretty much everything in this entire post is a gross oversimplification. The “real story” is in the paper itself - but of course it’s a lot more complicated, with lots of caveats and hedging. I’ve made the choice to gloss over a lot of important details to try to make all this easier to digest, but if you are a statistician or academic yourself, then I recommend you read the paper or at least post I did for Political Behavior's own blog, to get a sense of what I really did. Finally, it should go without saying that almost every other sentence in this post references work done by other researchers and scholars. I’m not going to bother to cite them all again here, since I do that in the paper itself. But, like all science, this work builds heavily on a lot of great work done by others.

 2 Bolding (but not italics) added. For what it’s worth, this quote is on page 116 of my dissertation: “Problems, Policies and “Paradoxes:” The Political Implications of American Concerns about Economic Inequality” but good luck finding a copy.

No comments:

Post a Comment