Out of Place

Ask a thousand Americans to rate their lives on a ladder from zero to ten, and the people who tell a phone interviewer they are lesbian, gay, bisexual or transgender sit, on average, about a third of a rung lower than everyone else. Ask them instead whether they felt worry, sadness or stress for much of yesterday, and the gap is far larger — they carry roughly half again as much daily distress. Those are two different measurements of the same lives, and they do not behave the same way as you move across the map.

This essay works from the Gallup US Daily survey, an enormous repository of American mood: across the years it ran the question, 62,193 respondents identified as LGBT, set against roughly 1.71 million who did not. The survey carries no design weight, so every number here is an unweighted sample estimate — a description of who answered, not a re-weighted portrait of the country. It is best read as a map of a very large sample, with the caveats that implies.

We keep two kinds of well-being strictly apart. One is evaluative — the cool, considered judgment of the Cantril ladder, and the share of people who count as Thriving. The other is experiential — whether worry, sadness and stress actually colored yesterday. They are never averaged into a single index, because the central finding of this piece is that they part ways.

01 · The national gapTwo measures, two stories

Start with the country as a whole. On the evaluative side the gap is real but modest. On the experiential side it is stark. The same group of people rates life only slightly lower than the average American, yet reports dramatically more of the day-to-day feelings that make a life hard to live in.

Read the columns the right way. The ladder gap of about a third of a point and the near-flat Thriving gap say that, on reflection, LGBT respondents place their lives only a little below others. The worry, sadness and stress gaps — each one large and well clear of its bootstrap interval — say that the texture of an ordinary day is much heavier. A life can be rated as roughly fine and still be felt as hard.

02 · The climate gradientOne curve bends, the other stays flat

Now let the map do some work. To each respondent we attach the climate of their county: how Republican it voted, how rural it is, how religiously dense it is. We fold the first two — the ones that move — together with the third into a single structural-stigma index, built by standardizing each measure and averaging them. Higher means a more hostile place to be openly LGBT, in the sense that the minority-stress literature uses the word.

Slide along that index and the two constructs split apart. The evaluative gap — the top panel — bends steadily away from zero: the further into hostile territory, the lower LGBT respondents rate their lives relative to their neighbors. The experiential gaps — the bottom panel — barely move. Worry, sadness and stress are elevated in the most welcoming counties and just as elevated in the most hostile ones. Daily distress is a constant; life evaluation is a slope.

The choice of axis matters, so the figure lets you swap it. Republican margin and rurality both produce the same widening evaluative slope. Religious density does not: split the country into thirds by the share of residents who are religious adherents and the ladder gap is essentially the same in each — it does not line up monotonically with the others. We keep religion in the composite index for honesty and transparency, but it carries little of the signal; the political and rural character of a place is what tracks the evaluative gap. The flat religion curve is not a failure of the data — it is a result, and one worth stating plainly.

03 · CompositionIs it just who lives where?

LGBT respondents are younger, better educated and more urban than the comparison group, and all of those traits move well-being on their own. So before reading the slope as a climate effect, we have to ask whether it is really a composition effect in disguise. We fit a regression of the ladder on an LGBT indicator, the stigma index, and their interaction, adding controls for age, sex, race, education, income and year, with standard errors clustered by county.

The interaction is the number that matters: does the gap widen with stigma once we hold composition fixed? It does. Unadjusted, each standard deviation of the stigma index deepens the gap by about of a ladder point; adjusted for who lives where, it still deepens it by about — an attenuation of roughly percent, but the slope survives, and it is statistically clear. A blunter cross-check agrees: compare LGBT and non-LGBT respondents inside the same age-by-sex-by-education-by-income cells, and the gap still grows from the friendliest third of counties to the most hostile. The geography is not an artifact of demographics.

04 · The disclosure floorWhy the widening gap is a lower bound

Here is the turn that should make a careful reader uneasy in a productive way. The measured gap depends on who is willing to be counted as LGBT — and that willingness is not constant across the map. Walk from the most Democratic decile of counties to the most Republican, and the share of respondents who tell the interviewer they are LGBT falls from about percent to about percent: a -fold drop.

Two mechanisms sit behind that decline, and this survey cannot separate them. One is concealment: in a less accepting place, fewer people disclose to a stranger on the phone. The other is composition: LGBT people move toward friendlier places, thinning the population that remains. Either way, the people still counted as LGBT in a hostile county are a selected group — more likely to be the resilient, the affluent, the urban, the willing-to-be-out. Measuring the well-being gap on that subset pulls it toward zero exactly where the true burden is probably largest.

Which sets up the most honest sentence in the piece. If the evaluative gap still widens as the climate worsens — even though it is being measured on an increasingly selected, increasingly favorable slice of the population — then the real gap in hostile places is at least that wide, and very likely wider. The widening curve is a lower bound. The disclosure line drawn across the gradient figure is what makes that argument visible: the slope and the selection point the same way.

05 · A locatorWhere the gap is widest

Finally, anchor the gradient in real places. The map shows the LGBT-minus-non-LGBT ladder gap by state, on a scale centered at zero. States with too few LGBT respondents to estimate stably — fewer than — are greyed out rather than guessed at. What remains is not a clean partisan map, because a state aggregates welcoming cities and hostile countryside together, but the widest gaps tend to fall where the hostile-county share is largest.

06 · What this shows — and doesn't

The honest ledger. Unweighted: these are sample means, not a re-weighted national portrait. One binary, on a phone: "LGBT" here is a single self-identification item, and it is disclosure-dependent — the people it captures shift with the climate, which is the whole point of the lower-bound argument. Association, not causation: county and state context are static place attributes joined to every year and read ecologically; controls attenuate composition but cannot rule out selection. The religion null: religious-adherent density alone does not track the evaluative gap, and we report that rather than bury it. The affect regime break: the yesterday-affect items changed sampling at 2013, so we never trend them across it and confine all affect figures to 2013–2017. And the two constructs stay separate from first panel to last, because merging a flat experiential gap with a sloping evaluative one would erase the finding.

What it does show is sturdy enough to say plainly. LGBT respondents carry markedly more daily worry, sadness and stress than other Americans, more or less wherever they live. Their considered judgment of their own lives is closer to everyone else's — but it slides further below their neighbors' the more hostile the surrounding climate, and it does so even after accounting for who lives where. And because fewer people are willing to be counted as LGBT in exactly those hostile places, that widening is the floor, not the ceiling.