Can Marriage and Church Explain It?

A companion piece established the basic fact: across fifty years of the General Social Survey, American conservatives call themselves “very happy” more often than liberals, by about seven points. The interesting question is why — and whether the gap is really about ideology at all, or just a stand-in for the fact that conservatives are more likely to be married, to go to church, to be older and more financially secure. Each of those is independently tied to happiness. So how much of the gap is left once you account for them?

The way to find out is to take a single sample of liberals and conservatives with complete records, predict who is “very happy,” and add the suspects one block at a time, watching the conservative advantage shrink. On this sample the raw gap is 7.0 points. Demographics — age, sex, race, region, survey year — trim it to 5.6. Marital status takes it to 4.1. Then churchgoing does the most work of all, dropping it to 1.6. Income and health change almost nothing. What remains at the end is 1.8 points — about a quarter of where we started.

Two things stand out. First, the explanation that does the heavy lifting is religion, not money. Adjusting for how often someone attends services cuts the gap by 2.5 points by itself — more than marriage and demographics combined did before it. Adjusting for household income, by contrast, moves the gap by essentially nothing: conservatives are not happier because they are richer. Second, the controls together account for about three-quarters of the gap — which means the popular claim that the conservative happiness edge is “just” marriage and church is mostly, but not entirely, right.

Because a quarter of it isn’t. After every adjustment, conservatives remain about 1.8 points more likely to call themselves very happy than otherwise-similar liberals. You can see the same residue without any model at all: split people by marital status and churchgoing into four corners, and conservatives stay ahead in every one — by 3.1 points among married non-attenders, 2.8 among the married faithful, even 1.5 among unmarried non-attenders.

What this shows — and what it doesn’t

This is an accounting exercise, not a verdict on cause. The controls are bundles of correlated traits, and some of them — marriage, churchgoing — may be partly downstream of ideology rather than rival explanations for it; if being conservative makes someone more likely to marry or attend services, then “explaining away” the gap through those variables understates ideology’s role. Adjustment also can’t see what isn’t measured: optimism, sense of meaning, trust, temperament. So the honest reading runs in both directions. The conservative happiness advantage is largely the well-documented happiness dividends of marriage and especially religious community — it is not a free-standing magic of ideology, and it is not about money. But it does not vanish when you account for those things. A modest, stubborn remainder — roughly a quarter, about 1.8 points — sits with ideology itself, or with something that travels too closely with it to separate in survey data.