The relationship between marriage, childbearing and the well-being of both mothers and children is complicated, as I've noted before. When, how and with whom women have children is linked to their educational attainment, their likelihood of employment, the academic prospects of their children and the chances that they'll live in poverty.

These relationships also travel in multiple directions: A single mother who has a child early in life is likely to have a harder time finishing her high-school education, a circumstance that will further hinder her job prospects and her future earnings. But we also know that non-marital births are more likely to occur among women living with disadvantage in the first place — in poverty, with less education — even before a child comes along.

The Census Bureau released a new report Tuesday that sheds a little more light on the connections between all of these data points: education, poverty, marriage and children. Census is publishing new data on women's fertility from the 2012 American Community Survey and the June 2012 Supplement to the Current Population Survey. Some of the trends will be familiar: The likelihood that a woman will be married when she has her first child in the U.S. has been declining, a trend that's paralleled by the rise of cohabitation:

Screen Shot 2014-07-08 at 3.29.42 PM

In the early 1990s, 70 percent of first-time mothers were married. More recently, that figure has dropped to 55 percent. The trend is even more noticeable among young married mothers — here under age 23:

Screen Shot 2014-07-08 at 3.31.11 PM

Since 2005, young women were more likely to be cohabitating or not married when they had their first child than married.

Most interesting, though, is an analysis that Census researchers Lindsay M. Monte and Renee R. Ellis conducted looking at the long-term prospects of those women above who had a first child out of wedlock. The June 2012 CPS Fertility Supplement asked women for the first time about their relationship status at the time of their first birth. And the answers make it possible to compare the well-being of women (and, by implication, their children) in 2012 with this earlier chapter in their lives.

The results of the regression analysis are summarized below, but in short: Women who weren't married when they had their first child — even if that event happened 10 or 15 years ago — were less likely than women who were married at first birth to have a high school diploma at the time of the 2012 CPS interview. They were more likely to be unemployed. They were less likely to be married today and more likely to be living in a blended family with stepchildren. The results were similar, but not quite as stark, for women who were cohabitating when they had a first child:

"Although other data suggest that women with a nonmarital first birth are likely to have lower educational attainment to begin with," Monte and Ellis write, "these data suggest that, even over the course of many years, women with a nonmarital first birth do not catch up to their counterparts whose first birth was in marriage."

The policy implications of what to do about this are not obvious: If women who have nonmarital first births also have less education, simply encouraging them to get married won't address other circumstances that contribute to — or hold back — their well-being and the future prospects for their children.