I'm not sure what to make of articles like this because, on the one hand, I appreciate someone trying to stay close to the data, but on the other hand, I don't

I'm not sure what to make of articles like this because, on the one hand, I appreciate someone trying to stay close to the data, but on the other hand, I don't trust the data they use, and feel like it could all be spun in different ways. Just to take one small example: Thompson cites as an example of happier workers a poll question showing that there was a 20% increase in people saying that it's "a good time to find a job.” But what does that mean? That there are lots of openings because people have in fact left their jobs they were unhappy with? Turnover means people are happier with their jobs? Or take this: 35% of people say "that opportunities abound for them to do 'more of what they do best.'" so the other 65% of people don't matter? Maybe it's unchanged since 2013 but is that a good benchmark? Are people thinking of other opportunities? Another elephant in the room is whether the happiest are happier, and the unhappy are less happy. This might or might not be the case, but talking about these survey respondents like they're a large homogenous group is problematic. It may be that there are areas of the economy that have in fact gotten much worse for large segments, even as they have gotten better for other segments. Focusing on the happy doesn't talk about the unhappiness of the unhappy, which is in fact what Thompson is trying to rebut. Finally, I've grown distrustful of these surveys in general. It's not clear to me that what people say in response to a pollster reflects anything meaningful about their actual behavior and emotional state in the way people want them to mean. I do not want to be dismissive because people should pay attention to them, but I'm not sure I've seen convincing evidence that aggregate survey responses track with societal trends in a way that suggests their unbounded validity. Even the figure tracking job satisfaction since 1987 — which I give credit to Thompson for posting — hides a lot. Sure, you can see the Great Recession in that figure, but if you didn't know about it, what would you make of it? What about the drug epidemic and declines in life expectancy? Should we be incorporating birth rate declines into the conversation? Are people unhappy with their jobs, or with some things about society in general? I bring that up not to say "life sucks for members of society right now" but rather, it's not always clear to me how you should use measures of things like "societal happiness" or what that even means. In some place as heterogeneous as the US, given all the problems we've had as a country, is it insensitive to have these kinds of discussions at all until we're more attentive to the worst off?