Does the Kid Next to the Doctor Really Become a Doctor?

Opening

Dear reader, you’ve probably heard the claim that “kids do better when they grow up in a good neighborhood.” School-district premiums, the educational effect of wealthy enclaves — it’s a conversation that plays out daily in real estate forums. But how would you actually prove this with data?

A team of American researchers really did it. They followed children who were 5 to 18 years old in the 1910 U.S. Census all the way to 1940 — a full 30 years. The sample: over 6 million people. And they asked a single question: “Did the occupation of the adult living right next door affect this child’s occupation 30 years later?”

The short answer: yes. And quite clearly, too. Today, let’s walk through what this paper uncovered, along with the mechanisms and limitations behind it.

What Can 100-Year-Old Data Actually Tell Us?

The study is formally titled “Like A Good Neighbor: Childhood Neighbors Influence Occupation Choice.” It was conducted by researchers at UMBC (University of Maryland, Baltimore County), Northwestern University, and Brigham Young University, and published in August 2025.

The researchers used original U.S. historical Census records. Back then, enumerators visited each household in person and recorded family members, occupations, ages, and more. Crucially, enumerators walked down a street in order as they recorded — which means the sequence of entries in the census roll actually reflects real physical adjacency between neighbors. The researchers used this sequencing to distinguish “next-door neighbor” from “someone further down the same street.” When validated, this method proved 91% accurate.

A natural question arises here: “Wait, this is a 2025 study — why use 100-year-old data?” There’s a reason. In the U.S., privacy regulations tightened starting in the 1940s, making this kind of detailed household-level data unavailable for public research going forward. This period happens to be the last dataset that allows for this kind of complete tracking analysis.

Of course, the limitations are real. 1910–1940 spans World War I and the Great Depression — a very particular historical moment — and women’s labor force participation was only around 30%. This historical context has to be factored in when interpreting the results.

The Doctor Next Door Changed My Career, 30 Years Later

Here’s the core finding of the study.

A boy who lived right next door to a doctor in 1910 was 41% more likely to become a doctor himself by 1940, compared to other children living on the same street. In absolute terms, that’s roughly a 0.3 percentage point difference — but given how small a share of the overall population becomes doctors, that’s a genuinely meaningful number.

For comparison: when a doctor is in the child’s own household (i.e., a parent is a doctor), the child is 12.4 times more likely to become a doctor. The next-door effect is roughly 1/30th the size of the family effect — but the fact that a neighbor with “not a drop of shared blood” produces this much influence is itself a striking finding.

The researchers extended the analysis beyond doctors to 50 major male occupational categories and 25 female categories. The results were consistent.

• Of the 50 male occupations, 46 showed a positive next-door effect
• Of these, 26 occupations showed a statistically significant¹ positive effect
• Not a single occupation showed a statistically significant negative effect

Pooling everything together, a boy who grew up right next door to an adult in a given occupation was, on average, about 10% more likely to enter that occupation himself. For girls, the figure was about 6% — a number that reflects the era’s low female labor force participation.

Why Does This Happen — Interaction, Mentorship, and the Physics of Distance

Just as important as the numbers is the question of “why?” The researchers propose several mechanisms.

First, the power of face-to-face interaction. The physical proximity of living right next door naturally creates opportunities for contact. A child observes the daily life of the adult next door, has conversations with them, and forms a concrete picture of what that occupation actually looks like. There’s a fascinating pattern the researchers found: when the next-door neighbor died before 1915 — meaning the child had less exposure time to that neighbor — the occupational transmission effect disappeared. That’s strong evidence that exposure duration matters.

Second, the effect fades as distance grows. Living right next door had a strong effect, and it persisted, if weakened, at two or three houses away. But beyond five houses, the effect essentially vanished. There’s a clear physical radius to a neighbor’s influence.

Third, homophily amplifies the effect. The occupational transmission effect was stronger when neighbors shared the same race, when the household had children of similar age, or when families shared the same birthplace. The frequency and depth of interaction determines the size of the effect.

Fourth, the effect was larger in high-income neighborhoods. According to the researchers’ analysis, neighbor-to-neighbor interaction was more active in high-income neighborhoods, while in low-income neighborhoods, even having a professional neighbor produced little effect because interaction itself was sparse. The same kind of neighbor can matter more or less depending on the density of the surrounding social network.

This Isn’t Just Occupational Copying — It’s About Income and Education

What the researchers found particularly notable is that the neighbor effect goes well beyond “kids simply copying the same job.”

Children who grew up next to neighbors in high-income, highly educated occupations (doctors, lawyers, etc.) had significantly higher education levels and income as adults. Conversely, children who grew up next to neighbors in low-income occupations actually showed a negative effect on their own income.

Even more interesting is a general human-capital spillover. For instance, a child who grew up next to a teacher wasn’t necessarily more likely to become a teacher — but was more likely to choose some other highly educated profession, like medicine or law. The neighbor wasn’t teaching a specific occupational skill; they were transmitting the value of education and an aspiration toward professional careers.

For girls, this effect carried particular weight. Girls exposed to neighbors working in occupations that were newly professionalizing at the time — teachers, nurses, stenographers² — showed significant increases in both education level and future income. Even within the era’s constraints, a neighbor as a role model opened a window onto new possibilities.

Oz’s Lens

Honestly, when I first saw this paper, my instinct was skepticism too: “What can we possibly learn from 100-year-old data?” But the more I read, the more I realized that this analysis is only possible because of data from this particular era.

Three things stand out to me about this research.

First, the power of “exposure” is far more concrete than we tend to assume. We often say vaguely that “environment matters.” But this study narrows that environment down to the unit of “the house right next door.” The fact that the effect disappears after just five houses tells us the key variable isn’t some abstract “neighborhood vibe” — it’s the actual frequency of face-to-face contact.

Harvard’s Raj Chetty and his team arrived at a similar conclusion. His research found that one of the strongest predictors of a child’s economic mobility is “cross-class social connection.” Simply occupying the same space doesn’t matter — what matters is whether people actually interact.

Second, this research offers a new angle on Korea’s school-district debate. In Korea, the standard for judging a “good neighborhood” is almost entirely about school quality. But what this paper is telling us is about neighbors, not schools. Even within the same school district, a child’s future can differ depending on who lives right next door. In Korea’s apartment-complex-centric residential environment, could the density of neighbor-to-neighbor interaction ever produce an effect like the one in this study? That’s a question without an answer yet.

Third, from the perspective of someone who works with data, the methodology here is genuinely meticulous. Turning the fact that census enumerators walked down streets recording households in sequence into something close to a “natural experiment” is a clever design choice. The additional validation using neighbor mortality data is impressive too. That said, I think we need to account for the historical peculiarity of this period. It’s hard to assume that the occupational choice patterns of a generation that lived through World War I and the Great Depression apply identically today.