Has Trust Gone Bust?

September 16, 2016
Posted by Jay Livingston

David Brooks was preaching again this week. His Tuesday column (here) was a jeremiad on the ills that “an avalanche of distrust” is bringing to US society.  (The phrase “avalanche of distrust” came probably from the headline writer, not Brooks. Brooks refers to “an avalanche of calumny.” Either way, the country’s being buried under a lot of bad snow.)

A generation ago about half of all Americans felt they could trust the people around them, but now less than a third think other people are trustworthy.

Young people are the most distrustful of all; only about 19 percent of millennials believe other people can be trusted. But across all age groups there is a rising culture of paranoia and conspiracy-mongering.

Brooks is partly right, partly misleading. He seems to be referring to the General Social Survey, which, since 1972, has asked regularly about trust.  The GSS has three items that pertain to trust.
  • TRUST – Generally speaking, would you say that most people can be trusted or that you can't be too careful in life?
  • HELPFUL – Would you say that most of the time people try to be helpful, or that they are mostly just looking out for themselves?
  • FAIR – Do you think most people would try to take advantage of you if they got a chance, or would they try to be fair?
The GSS data does show some decline in all these since 1990 – i.e., a generation ago.

(Click on an image for a larger view.)
Of the three variables, TRUST has declined the most. The percentage of people saying that most people could be trusted fell from 39% to 30%, Brooks’s statement about “half of all Americans” a generation ago being trusting is a bit misleading as it implies that 50% was standard year in year out. In fact, in only one year, 1984, did the percentage reach that level. As for the other variables relevant to the “culture of paranoia,” perceptions of other people’s helpfulness also declined; perceptions of their fairness changed little.

What about the age differences Brooks notes?  I extracted GSS data on the Trust variable at three different periods – 1972-1976, 1989-1991, and the most recent years that we have data for.

In every period, the young are the least trusting, but the difference between them and people in older age groups is much greater now than it was 25 or 40 years ago. That’s because millennials, as Brooks correctly notes, are much less trusting than were their 18-30 year old counterparts in the 1970s.

But what about the rest of us? According to Brooks, “across all age groups there is a rising culture of paranoia.”  In the 1972-76 period, in the youngest group, 38% were trusting. In 1990, those people would be in their early 40s to early 50s. In that year, that age group was somewhat more trusting than they had been 25 years earlier. And 20 years later, when they were in their late 50s and up, they were still as trusting as they had been before. The same is true of the people who were 30 and up in the 1970s. Similarly, the youngest group in 1990 had the same level of trust twenty years later – about 30%.

Trust seems to be remarkably resilient – impervious to the vicissitudes of aging or of social and political changes. The Watergate era, the Reagan years, the terrorist attacks of 9/11, the recent recession – none of these seems to have made much difference in each cohort’s level of trust.  Go figure.

Brooks’s sermon then turns from the aggregate data to the effect on people’s souls.

The true thing about distrust, in politics and in life generally, is that it is self-destructive. Distrustful people end up isolating themselves, alienating others and corroding their inner natures.
Over the past few decades, the decline in social trust has correlated to an epidemic of loneliness. In 1985, 10 percent of Americans said they had no close friend with whom they could discuss important matters. By 2004, 25 percent had no such friend.

That finding, which made headlines a decade ago, has since been questioned if not debunked.The GSS data it’s based on contained a coding error, and other surveys have found no such drastic increase in friendlessness. Claude Fischer has an excellent blog post about this issue (here). He includes this graph based on results from the Gallup poll.

Brooks obviously is not interested in these corrections. He is, after all, crying “avalanche” in a crowded political theater. But it tuns out there’s not all that much snow. The change to worry about is that over the last 25 years, each new cohort is less trusting, and this is one time when hand-wringing about what’s wrong with kids today might be appropriate. Those attitudes, once formed, are little effected even by major changes in the society and government.


Andrew Gelman said...

I don't trust David Brooks because there are multiple times that he's spread erroneous data, and when people call him on it, he just expresses annoyance, he never either admits error or changes the inferences he claims to have made based on his fake numbers.

Jay Livingston said...

Yes, I was thinking of your posts about him when I wrote that line about Brooks not being interested in corrections.

Ron Skurat said...

Brooks isn't a serious commentator, he's a performer. Not quite as obviously as Maureen Dowd is, but he's producing a cultural product for middlebrow entertainment. The problem is that government, corporations, media, and other institutions are run by the middlebrow.