CDC Data and Dating: What 50 Years of Government Research Actually Reveals

Most people associate the CDC with infectious disease and vaccination campaigns. That is fair. But buried inside the CDC's research portfolio is one of the most rigorous, longitudinal datasets on American relationship behavior ever assembled: the National Survey of Family Growth, running continuously since 1973. It covers marriage, cohabitation, divorce, fertility, and family formation across every major demographic group. It is publicly available. And almost no one who is actually navigating the dating market has ever looked at it.

Dating advice, by contrast, tends to draw from far weaker sources: opinion, anecdote, social media trends, and companies whose revenue depends on keeping you single and scrolling. The NSFG is none of those things. It is epidemiological data on how relationships actually form, persist, and dissolve in the United States. That makes it genuinely useful. This article extracts the findings most relevant to women who are serious about finding long-term partnership.

What the NSFG Actually Tracks

The National Survey of Family Growth is a nationally representative, face-to-face interview study of adults aged 15 to 49. It has been administered in multiple waves since its launch and is updated regularly to capture how family formation behavior changes over time. The survey collects detailed information on relationship history, cohabitation, marriage, divorce, sexual behavior, and fertility.

The critical distinction: the NSFG reflects the actual adult population of the United States. It is not a self-selected sample. It is not shaped by a commercial agenda. Unlike surveys run by dating apps (which capture only people who use those platforms, and which are conducted with engagement in mind), the NSFG is designed to produce accurate population-level data. When it tells you something about how relationships work, it is drawing on a nationally representative sample with decades of longitudinal follow-up.

Cohabitation No Longer Signals Anything About Direction

One of the sharpest changes in the NSFG data over five decades is the normalization of cohabitation before marriage. In the 1960s and 1970s, most couples who married had not lived together first. Today, the large majority of first marriages in the United States are preceded by cohabitation. Living together has gone from exceptional to near-universal.

Here is what that shift means for you: cohabitation itself is no longer a useful signal about commitment trajectory. The man who agrees to move in with you is not telling you much. The relevant questions are about what that step means to each of you, specifically. Is the move toward cohabitation a deliberate step toward marriage? Do both of you understand it that way? Has anyone said the word "marriage" out loud?

NSFG data on cohabitation dissolution makes this point sharply. A large share of cohabiting relationships that do not convert to marriage within a few years end in dissolution rather than marriage. Sliding into shared housing does not naturally lead to commitment. Explicit conversation about direction is what leads to commitment.

Second Marriages Fail at Higher Rates Than First Ones

NSFG and related CDC data consistently show that second marriages dissolve at higher rates than first marriages. This surprises most people. You would expect someone who has been through one marriage ending to bring more wisdom, more selectivity, and clearer priorities to a second one.

Two things explain the gap. First, people who have survived one divorce know that leaving is survivable. They are more willing to exit a second marriage that is not working. Second, the behavioral patterns that contributed to the first marriage ending tend to persist into subsequent relationships unless they are explicitly identified and addressed. Most people do not do that work.

For dating purposes: a man's divorce history is relevant, but it is not disqualifying on its own. The more useful question is what he actually learned from the experience. A man who has done real work on the patterns that ended his marriage can be an excellent partner. A man who frames his divorce entirely in terms of his ex-wife's failures, and who has changed nothing about his own approach, is on track to produce the same outcome again.

Education Matching Beats Education Level

CDC data has consistently shown that education level correlates with marriage rates and stability, but the relationship is not what most people assume. The stronger predictor is not how educated either partner is in absolute terms. It is how closely matched the two partners are in educational background.

Couples with similar educational levels show higher long-term stability than couples with large educational disparities, even after controlling for income and other socioeconomic factors. The likely mechanism is alignment of values, lifestyle expectations, and social world, all of which correlate with educational background more reliably than education does as a standalone credential.

The practical implication: filtering for a specific degree as a proxy for intelligence or ambition may be less productive than paying attention to whether a man's actual values, lifestyle, and expectations align with yours. Those things matter more than the credential on the wall.

Age at First Marriage Predicts Divorce Risk

One of the most robust findings across multiple waves of CDC data is the relationship between age at first marriage and the probability of divorce. Marriages that begin before age 25 show substantially higher dissolution rates than those beginning in the late twenties or early thirties. The risk does not increase significantly for marriages beginning in the early-to-mid thirties compared to the late twenties.

For evaluating a man's dating history, this finding changes the interpretation of a divorce. A man who married at 22 and divorced at 26 has a different profile than one who married at 34 and divorced at 39. Early marriages that end often reflect decisions made before either person had the self-knowledge or life experience to choose well. That does not make the subsequent person a stronger partner automatically. But it changes what that history means, and it warrants different questions.

Past Relationship Behavior Is Your Best Predictor

This is the finding with the most direct practical application, and it is the most consistently ignored. Across relationship types and demographic groups, NSFG data shows that past behavior in intimate relationships is the single strongest predictor of future behavior. Men who have sustained long-term functional relationships are more likely to sustain them again. Men with a consistent pattern of short relationships ending in conflict or avoidance are more likely to reproduce that pattern.

Most people treat relationship history as a secondary topic, something to learn about naturally over time. The data says it should be a primary screening criterion. Not as a punishment for past failures, but as the most reliable signal available about what a person is actually capable of in a relationship.

• 1

  Ask About the Longest Relationship Early

  What was it, how long did it last, and how did it end? This single question reveals more about long-term compatibility than almost anything else you will learn in early dating.

• 2

  Listen for Accountability vs. Blame

  A man who can describe his own role in why a relationship ended, without deflecting entirely onto the other person, is demonstrating self-awareness. That is a stronger signal than relationship history alone.

• 3

  Pattern Over Instance

  One short relationship that ended badly is not informative. A sequence of short relationships that all ended the same way is a pattern. Patterns repeat.

What Population Data Cannot Do

Population-level data improves your initial screening and calibration. It does not determine anything definitive about any specific individual. A man who fits the statistical profile of a poor long-term bet can become an excellent partner. A man who checks every statistical box can still be deeply incompatible with a specific person.

The value of data in dating is efficiency at the early stages. It prevents the most common and predictable mistakes. It helps you arrive at the relationship phase with more of the right people and fewer of the wrong ones. Chemistry and connection still have to develop in person. The data just improves the conditions under which that happens.

How the Calculator Connects This Data to Your Decisions

The Online Dating Calculator was built to make NSFG data accessible without requiring you to read survey methodology papers or interpret epidemiological tables. The calculator translates the most relevant population statistics into a clear picture: what percentage of men in the actual dating market meet the specific criteria that predict long-term compatibility.

The filters draw from the variables with the strongest longitudinal support: relationship history, age, children status, life-stage alignment, and others the CDC has tracked across multiple waves of data collection. The output is not a ranked list of acceptable men. It is a calibration tool. Seeing that your specific combination of criteria applies to a particular share of the male population tells you where the real constraints in your search are, which criteria create the largest filtering effects, and where the most productive trade-offs exist if you choose to adjust your priorities.

This is information that previously required knowing where to look and how to interpret what you found. The calculator exists so you do not have to do that work from scratch.