Monday, July 21, 2014

The Atlantic's View of Polls Doesn't Make Any Sense

Photo Credit: David Shankbone via Wikimedia Commons

The Atlantic's Derek Thompson provides us with the predictably provocative headline, "Millennials' Political Views Don't Make Any Sense". In case we're easily offended, the tag line reassures: "That's not a harsh assessment. It's just a fair description."

This would be hardly surprising news. Most people aren't political analysts, and many are only casually interested in politics. Worse still, you could probably show that even most politicians have views that are not totally coherent, artfully crafted as they may be. (Barack Obama's pre-2012 views on gay marriage never had a semblance of logical consistency.) And as the youngest voting block, it wouldn't be surprising that the under-30-year-olds would have a less than stellar grasp of the major issues facing society.

So I was actually pretty surprised to find that Thompson doesn't prove his case at all. Not even close. I'll take the evidence he offers in turn.
[Millennials are] for smaller government, unless budget cuts scratch a program they've heard of. They'd like Washington to fix everything, just so long as it doesn't run anything.
These are obviously broad summaries, and come from a poll done by the Reason Foundation, hardly an unassailable source. Even still, as broad summaries, the views aren't totally incoherent. It's reasonable to say that you think the government should be smaller, but it should preserve programs that you're familiar with. And it's not a crazy idea that Washington could offer a lot of solutions to various problems without taking over anything in particular.

Consider next what Thompson calls a "smorgasbord of paradoxes":
  • Millennials hate the political parties more than everyone else, but they have the highest opinion of Congress.
  • Young people are the most likely to be single parents and the least likely to approve of single parenthood.
  • Young people voted overwhelmingly for Obama when he promised universal health care, but they oppose his universal health care law as much as the rest of the country ... even though they still pledge high support for universal health care. (Like other groups, but more so: They seem allergic to the term Obamacare.)
Not one of these is a paradox. You could think that political parties are a really bad force in politics, but that the individuals who make up the congress are on a whole doing better than the country seems to think. You could think that single parenthood is bad, but still end up as a single parent; more plausibly, the group of individuals who are the single parents could be quite distinct from the group who think single parenthood is bad. And you can support universal healthcare, but not think that Obamacare is the best way to achieve it. This last view is not only coherent, but it also seems to be the correct position to hold.

In fact, Thompson acknowledges that some of these views can be coherent. He writes:
... you can technically support (a) reducing the overall tax burden and (b) raising taxes on the wealthy by raising the investment tax and absolving the bottom 50 percent of Social Security taxes. Somehow, I think what's happening is simpler than young people doing the long math of effective tax rates. I think they're just confused.
You don't need to do math to see that reducing overall taxes and raising the portion paid by the rich is possible. It's possible on its face. Instead of realizing this, Thompson informs us that it's already his opinion that they're confused. I would be interested in whatever led him to that conclusion, rather than this series of pseudo paradoxes.

This is also strange:
Perhaps it should be [confusing] when we're using a couple thousand subjects to guess the collective opinions of 86 million people. 
As far as I know, that's generally how polling works. As far as sample sizes go, 2000 is pretty impressive. As long as they were selected with sufficient randomness, there's good reason to think the poll is representative.
Millennials are more liberal than the rest of the country, particularly on social issues, but they get more economically conservative when they make more money. 
This is not surprise, nor is it divergent from larger trends. It is also not incoherent.
  • On spending:
    Conservatives can say: 65 percent of Millennials would like to cut spending.
    Liberals can say: 62 percent would like to spend more on infrastructure and jobs.
  • On taxes:
    Conservatives can say: 58 percent of Millennials want to cut taxes overall.
    Liberals can say: 66 percent want to raise taxes on the wealthy.
  • On government's role in our lives:
    Conservatives can say: 66 percent of Millennials say that "when something is funded by the government, it is usually inefficient and wasteful."
    Liberals can say: More than two-thirds think the government should guarantee food, shelter, and a living wage.
  • On government size:
    Conservatives can say: 57 percent want smaller government with fewer services (if you mention the magic word "taxes").
    Liberals can say: 54 percent want larger government with more services (if you don't mention "taxes").
Let's take this in turn.

On spending: You can cut total spending and spend more on infrastructure and jobs.

On taxes: You can cut taxes overall and raise taxes on the wealthy.

On government's role in our lives: You can think that the government is often inefficient and wasteful, but that we still need it to ensure a certain baseline welfare standard. Even if a service is provided inefficiently, it's likely still better that it's provided at all.

On government size: Now, there are ways to expand the government without raising taxes. You could simply grow the economy (and thus revenue), or raise deficit spending for a while.  However, I acknowledge that these postulations would be a stretch to make these views look consistent.

Fortunately, there's no need to make these views consistent. If 57% of millennials believe in smaller government when taxes are mentioned, and 54% believe in bigger government without taxes being mentioned (and we assume that bigger government means higher taxes), then only an overlapping 11% of people need to have incoherent views (probably a bit more, given that some people decline to answer). But if 11% of millennials have incoherent views, does that justify the assertion that millennials' political views make no sense? Of course not. Only a minority of them demonstrably have views that make no sense.

And once we realize this, the entire argument looks silly. Even if a majority can be found to assent to each of two conflicting views, this only implies that a small minority have irrational views. 

As I said, I wouldn't be surprised if many millennials had a very confused political views. I would be surprised, rather, if they didn't. But I think this would hold across generations, and is to some extent a mundane fact. It's quite fashionable to complain and criticize millennials, but little of the work that does bothers to actually compare them to any other group.

1 comment: