People Aren't Better Off Than Income Trends Show

Matt O'Brien at Wonkblog recently picked up a month old poll that shows the majority of economists saying that people are better off now than income trends suggest. The argument here is that price indexes have failed to fully account for the change in the quality of goods and therefore understate how much true inflation-adjusted income has risen over time.

Insofar as we cannot tap directly into people's brains and measure their hedonic activity, this debate is inevitably driven by people's feelings and intuitions about how much things are better now than before. One way to try to prime these intuitions is to ask, as O'Brien does, this kind of hypothetical:

Try this thought experiment. Adjusted for inflation, would you rather make $50,000 in today's world or $100,000 in 1980's? In other words, is an extra $50,000 enough to get you to give up the internet and TV and computer that you have now? The answer isn't obvious. And if $100,000 isn't enough, what would be? $200,000? More? This might be the best way to get a sense of how much better technology has made our lives—not to mention the fact that people are living longer—the past 35 years, but the problem is it's particular to you and your tastes. It's not easy to generalize.

When people embark on this sort of priming, they always ask people to hypothetically imagine themselves living in a world three or four decades ago, and then ask them what price would be required to get them to do so. Since you are attached to the technologies and media of your time, you inevitably say that it would require a huge premium to go back in time.

But this is a very strange approach to this question. The reason it is strange is that we actually have people who are alive right now who were in their adulthoods in 1980. These people know what the technology was like then and now. And, at present, they have some degree of control over which set of technologies they use.

Using smartphone adoption as a proxy for these people's technological preferences, it's clear that the people who actually lived as adults through both technological periods overwhelmingly prefer older technologies:

The super-majority of people over the age of 55 do not have a smartphone. Additionally, a good chunk of those over that age that do have a smartphone don't really use it like a smartphone (instead they treat it more like an older phone):

Judging from these people's preferences, you'd have to conclude that, in fact, older technologies are preferable to newer technologies. You don't need a hypothetical to determine whether living in the past was better: these are people who lived in the past and the present and clearly prefer the way they lived in the past, at least when it comes to the technologies that are supposed to have made life dramatically better (as incomes stagnated).

In saying all of this, I realize there is a semi-obvious rebuttal: older people only prefer older technologies because they grew up with them, were habituated to them, and feel familiar to them. This is a very compelling rebuttal, but it ends up swallowing the entire argument that people are better off than income trends show. Old people prefer older technologies because they are habituated to them, but the same is also true for younger people and newer technologies.

People don't want to adopt very radically different technologies than the ones they are familiar with, and may even be willing to pay a huge premium to avoid the pain of going through that transition. But, since this willingness to pay a premium goes both ways (olds don't like new technologies, and youngs don't like older technologies), it cannot be used to show any one set of technologies is actually delivering more hedonic value than the other. Put simply: the premium you are willing to pay for the technologies you like is, in major part, a premium for familiarity, not a premium for superiority.

This is not to say that some technology isn't better and more additive to well-being than others. It's just to say that the intuition pump of asking people habituated to current technology to consider what it would be like to live under older technology is not really helpful in proving anything. It doesn't disentangle better from familiar, and therefore does not properly identify how much each factor is actually driving any willingness to take an income cut to stick with what you currently like.