The most rigorous longitudinal data on marriage outcomes — satisfaction across years, divorce rates by cohort, the patterns of separation and reconciliation across the life-course — tells a more hopeful story than the cultural commentary about marriage tends to suggest. The divergence between the data and the commentary is worth attending to, on its own terms.

What the data shows

The data shows, on the most consistent set of recent studies, that marriages contracted in the past fifteen years are showing better long-arc outcomes on most measurable dimensions than marriages contracted across the preceding several decades. The improvement is not transformative; it is real.

The improvement reflects the working-out of patterns that had been visible in the underlying data for some time. People are getting married later, on average, with more careful selection of partners; the resulting cohorts are more durable.

What is more variable

What varies more than the broader pattern would suggest is the geographic and demographic distribution of the outcomes. The improvements are not evenly distributed across the population; specific populations are showing meaningfully better results than others.

The reasons for the variation are partly economic, partly cultural, partly the consequence of access to support structures that some populations have stronger than others. The variation is part of the longer-arc story; it does not change the headline pattern.

What the commentary keeps missing

The commentary keeps missing the basic fact that the data, taken in aggregate, is more positive than the framing of much marriage discourse implies. The framing tends to lean on selection effects in the cases that produce media coverage; the broader population is operating differently.

The longer view

The longer view is that marriage as a social institution has, with notable consistency, adapted to the changing structures of the broader society. The adaptation has been imperfect and uneven. The aggregate result has been more positive than the cultural conversation has tended to suggest.

What this is not saying

The data is not saying that everyone who wants a successful marriage will produce one through good intentions alone. The structural questions remain real. What the data is saying is that, for the people who navigate the structural questions with care, the outcomes have been improving rather than deteriorating.