There is an image, somewhat faded now, of how in the course of the 20th century the rough equality of the sexes was achieved. Starting in the years before 1900, a relative handful of women went to college. Then as suffragettes, their children marched and won the right to vote.  And all the rest followed in due course.

It's a caricature that avoids many of the most salient facts, according to Harvard University professor Claudia Goldin, who gave the sole plenary lecture at the annual meeting of the American Economic Association in Boston last week. Her Ely Lecture is available here.

To focus on those boisterous and hopeful early years was to miss where the real action took place -- in labor markets, she said. The transformation unfolded in four distinct phases over a century -- three evolutionary developments followed by a "quiet revolution" for which Goldin, herself a 1967 college graduate, has had a front-row seat.

That revolution, a profoundly sudden and discontinuous change occurring from the 1970s to the present day, was accomplished by the generation born in the late 1940s. Many of them were unaware that they were part of a grand transformation -- "unwitting foot soldiers of an upheaval that would alter women's employment, education and family," said Goldin.

But would the transformation they had wrought be permanent? Some were worried now that the redefinition of woman's role in society was poised to take a step backwards, Goldin said, in the form of a new choice available to women sometimes described as "opt out."  

For after rising without interruption for at least a century, labor market participation rates for women of almost all ages, education levels and marital statuses had leveled off since 1990 in the vicinity of 75 percent.

Would the plateau turn out to be permanent?  Was there a "natural rate" of working women, perhaps somehow rooted in their role in childbearing?  Or would their collective participation in the labor force at some point resume its steady climb?

Maybe, said Goldin. Maybe not. Certainly time would tell. But , more important, only a thorough understanding of the past would permit clarity in the future.

It was women's understanding of themselves that had changed. Goldin grounded her distinction between evolution and revolution in three particular aspects of female self-understanding.

The first aspect, she said, had to do with a woman's "horizon" -- with whether she expected that her participation in the workforce would be long and continuous, or intermittent and brief.

The second had to do with her "identity" -- with the degree to which a woman finds individuality in her job, occupation, profession or career, as well as in her family.

The third had to do with her "decision-making" -- whether her decisions about her work were made fully jointly, if a woman is married or in a long-term relationship; or whether she viewed herself as a secondary worker, who would do the best she could after taking her partner's decisions as given.

All three aspects had changed greatly over time.

During the first phase of the modern evolution, which lasted from 1890 until 1930, the first generations of college-educated women made their appearance in the workforce, mostly as teachers and clerical employees. But the vast majority of woman workers were poorly educated, often from immigrant families. Manufacturing jobs were dirty, dangerous and repetitive; domestic service offered long hours with few days off.  Working women almost always quit when they got married.  Employment rose to 27 percent in 1930 for females 25 to 44 years old, from around 18 percent in 1890; to 10 percent for married white women 35 to 44 years of age from around 3 percent at the beginning of the period.

The second phase lasted from 1930 to 1950. Typewriters and telephones meant plenty of new jobs. The nation's new high schools graduated plenty of willing workers. "Marriage bars" insured that many women left their jobs as soon as they got married. But the Depression and World War II meant that a much greater number of married women went to work for one reason or another.  As their husband's wages went up, they tended to stay home.  But the appearance of labor-saving appliances such as refrigerators and washing machines made it easier to choose to work outside the home.

The third phase, from the 1950s to the 1970s, provided the roots of the revolution. There was much greater acceptance of women in the workplace in the years after World War II with the rising demand for labor, even though their claim on many kinds of jobs they had performed during the war disappeared. Marriage bars vanished; part-time jobs for teachers, secretaries, librarians, nurses proliferated. Soon the average working woman was more educated than the average woman at large, reversing the previous situation. But a typical interview question, even for college graduates, was "How well do you type?" Women were still mostly secondary earners.  College was a way to meet spouses.  Few women who married expected to work most of their lives. But that was before the advent of the birth control pill, and the slackening of strictures against divorce, dramatically changed the rules of the game. Some 46 percent of married women in the 35-44 year old age group were working by 1970, up from 25 percent in 1950.

It was only when the generation of the baby boom came of age that Goldin's fourth phase emerged, her "quiet revolution." This took the form of sudden and dramatic changes in those three attributes of the self -- horizon, identity and decision-making. Teen-aged women in the mid-1960s began to recognize that their lives would differ significantly from the lives of their mothers. They expected to work longer, so they took more science and math in high school, invested more heavily in their college educations and applied to professional schools. They married later -- median age at first marriage rose an astonishing 2.5 years in the early 1970s (for those born between 1950 and 1957).  By the mid-1980s, the median college graduate was marrying at 25. Occupational choices shifted; from a trickle of admissions in the early 1970s, women by the mid-1990s were being admitted in equal numbers to schools of science, medicine dentistry, business and law. Their earnings increased relative to those of men. The horizons of their working lives had lengthened dramatically.  Instead of jobs they now had careers.

But decision-making apparently is still a problem. Goldin didn't even try to trace developments in this most hazardous terrain.  

It was in the early 1990s that female labor-force participation fetched up on its current plateau -- 76 percent for married college graduates in their thirties, 72 percent for married high school graduates. The participation rate for mothers of infants probably has declined somewhat. Perhaps a "natural rate" of workforce participation had been reached in the United States, somewhat lower than if day care for children was an obligation of the state, as it is in Scandinavian countries. Is it real?  Will it last?  Are significant numbers of women rethinking the question of "balance" in their lives? Significant numbers of men?  

It is, said Goldin, simply too soon to tell. The available evidence does not indicate that the "opt out" pattern has increased, that a new age of the "stay-at-home mom" is at hand.  An intensive study of individuals from the class of 1980 at 34 selective colleges and universities completed in the mid-1990s showed that the median woman was never out of work for more than six months. For women with children the sum of all out-of-work spells was barely two years. But it will be a good long time before we have a clear statistical idea of how patterns of work and marriage are shifting in the present day, or whether, after a century of rapid change, a stable equilibrium has been reached.