I Trust My Fellow New Yorkers

Social cohesion has been slipping away in suburban America, but not in the big city.

We're all in this together.

Photographer: Don Emmert/AFP/Getty Images

The New York Times ran an article last week about jury selection for the high-profile case known (in the pages of the New York Post, at least) as the "killer nanny" trial, and how it showed Manhattan to be riven by divides of class and wealth. A sample:

Whether potential jurors had a nanny or not, either as a parent or a child, whether they worked as one, and what that experience was like, has become part of the questioning. And the voir dire process itself has become a kind of lens onto the different New Yorks -- one inhabited by those who can afford nannies, the other populated by those who cannot, or who might have worked as a nanny themselves, like potential juror No. 18, who spent a summer caring for her nieces and nephews as a full-time job. The varying attitudes of the city are on stark display within the wood-paneled courtroom, as are searing emotions.

None of that is incorrect, but as someone who spent most of last week sitting in that courtroom waiting anxiously for my name to be called (it never was), I have to say that I was left with a different impression. Jury duty is one of the great levelers in Manhattan. It's probably a great leveler elsewhere, too, but I only have personal experience of it here, where there are certainly some big economic and cultural disparities to be leveled. This was my third go-round on jury duty -- on each of the previous two, I was selected for a short criminal trial -- and once again I departed with feelings of increased admiration for and connection to my fellow New Yorkers.

Yes, there was some whining and dubious-excuse-making by people looking to avoid getting stuck on a 15-week trial, as the Daily News reported, but can you blame them? 1 For the most part, the would-be jurors questioned by Judge Gregory Carro and the attorneys for the prosecution and defense came across as decent, thoughtful people of widely varying backgrounds and economic circumstances wrestling honestly with the difficult question of whether they were capable of fairly determining whether someone was mentally competent to be found guilty of the murder of two young children. (Defendant Yoselyn Ortega doesn't seem to be disputing that she killed them.)

In her farewell column for Bloomberg View last week, Megan McArdle wrote of a recent reporting trip to Denmark and the high levels of trust and social cohesion that make the Danish system work. The U.S. is much bigger and more diverse than Denmark, and it scores lower in trust surveys than the small nations of Northern Europe (and one giant nation in Asia: China). 2 The percentage of Americans responding in the affirmative to the survey question "Can people be trusted?" has also dropped a lot since the 1980s. 3

Can People Be Trusted?

Percentage of Americans answering yes

Source: General Social Survey

Dotted lines represent years with no data.

As I shuttled back and forth to jury duty, though, I started wondering whether New York City might be bucking that trend. The city is definitely not homogeneous -- 37 percent of its residents are foreign-born, 49 percent speak a language other than English at home, 52 percent were born outside New York state. The racial/ethnic breakdown is 32 percent non-Hispanic white, 29 percent Hispanic, 22 percent non-Hispanic black, and 14 percent Asian (which leaves another 3 percent or so who self-identify as some kind of mix). It is home to some of the country's richest people, but one-fifth of its residents are below the poverty line. Manhattan is somewhat whiter, richer and lower on immigrants than the rest of the city, but it has a higher percentage of residents born out of state.

New York City isn't all that much more populous than Denmark (8.5 million to 5.7 million), though. It's also pretty compact, and for all our differences, we New Yorkers do have a lot of shared experiences. Most of us ride the subway and/or bus, all of us walk on the sidewalks and share the parks. We are in each other's faces in a way that people who live in car-oriented suburbs often aren't. That constant proximity of millions of other people lends an anonymity that can be both freeing and isolating -- the 46 percent of New York County (aka Manhattan) households that are individuals living alone is a higher share than in any other county in the country. But the other boroughs are near or below the national average on that metric, and I can attest that raising a kid (and a dog) in Manhattan entangles one in the same networks of connections and friendships and civic commitments that it does anywhere else. New York City has also been on a mostly unbroken rise since the early 1990s: Crime has fallen precipitously, while employment is way up. Labor-force participation has also risen, in sharp contrast to the downward national trend.

So are New Yorkers figuring out how to trust one another? Well, sort of. Thanks to a 2000 paper on "Measuring Trust" by economists Edward Glaeser, David Laibson and Jose Scheinkman and psychologist Christine Soutter -- which found that answers to "Can people be trusted?" are in fact somewhat predictive of actual behavior 4  -- I learned that answers to the General Social Survey, 5 in which that question has been asked since 1972, can be sorted by the population of the place where the respondent lives. Which I then went and did:

Trust in Big Cities and Small Ones

Percentage answering that people can be trusted, by population of place of residence

Source: General Social Survey

Dotted lines represent years with no data.

It looks like trust has held steady(ish) among people in big cities since the early 1990s while declining everywhere else. To make the contrast a little clearer, I combined all the responses from 1972 to 1993, and all those from 1994 to 2016. Not only does this divide things pretty much in half chronologically, but 1994 also marked the beginning of the great crime decline in urban America. Because I was combining surveys from lots of years and thus increasing the sample size, this time I also created a category of cities of more than 5 million people -- which in the U.S. contains only New York City. Here's what I got:

Trust Has Declined Less in the Big City

Percentage answering that people can be trusted, by population of place of residence

Source: General Social Survey

*Respondents whose place of residence has a population of more than 5 million.

So New Yorkers are less trusting than other Americans, but the gap is a lot smaller than it used to be. The great trust decline that has beset the rest of the nation in recent decades is not apparent here -- or maybe it just happened earlier. The raw data even shows a rising trust trend in New York City since the mid-1990s. The sample sizes are so small (the number of respondents from places with more than 5 million inhabitants has ranged from 17 to 93 per year) that I wouldn't make too much of it. But it does fit with my personal experience.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
  1. The judge had tried to avoid this by having all prospective jurors fill out questionnaires about whether serving on a trial for almost four months would cause insurmountable difficulties, which eliminated 70 to 80 percent of the original pool, but some people hadn't heard what he said about the length of the trial, and others clearly had second thoughts when they learned what the case was.

  2. As my Bloomberg View colleague Noah Smith notes in a column today, the U.S. also scores higher than a bunch of smaller, less-diverse countries.

  3. The link to the "Can people be trusted?" survey has been acting weirdly in Chrome browsers. Click on it and you're likely to get a message telling you that "The page you were looking for doesn't exist." This is a lie. All you have to do, in fact, is copy the URL of the supposedly non-existent page and paste it over itself (or in the address bar of any other browser tab or window), and it should work fine. Or, if you'd prefer, just search on "trust GSS" and that will probably get you there.

  4. "In summary, to determine whether someone is trusting, ask him about specific instances of past trusting behaviors. To determine whether someone is trustworthy, ask him if he trusts others."

  5. Conducted by the organization that now goes by the unwieldy name NORC at the University of Chicago. NORC used to stand for National Opinion Research Center.

To contact the author of this story:
Justin Fox at

To contact the editor responsible for this story:
Brooke Sample at

Before it's here, it's on the Bloomberg Terminal.