Harvard's conception of merit is subjective.

The irony of this situation cannot be missed. After all, Harvard invented meritocracy.

Journalist Nicholas B. Lemann '76, in his book The Big Test: The Secret History of American Meritocracy, credits University President James B. Conant '14 with founding the concept.

"Conant wanted a system for selecting an administrator class," Lemann says, "of picking out a few people with high IQs and training them to become leaders of society."

The result was that the College tended to accept high academic achievers whose test scores were a good predictor of their success at the university level. Other colleges followed suit and, by the mid-1960s, most selective academic institutions used the Conant model for their admissions.

But student activists and educational experts were beginning to realize the effects poverty had on achievement and on how socioeconomic status correlated with the credentials one brought to the table.

This correlation showed. The typical Harvard student was, even in that age of social foment, a white male who had been educated in a Northeastern prep or private school.

So though Harvard had always admitted more minority students than other Ivies, it began to reexamine its philosophy in the late '60s and early '70s.

In the early '70s, the school re-defined its educational mission. In contrast to Conant's model, which had promoted a cognitive elite, the new system promoted and privileged diversity. The goal of the College was to admit a diverse class--one in which students would educate each other. Taking account of a student's race and gender made sense.

The Supreme Court, in its famed Bakke v. California decision, cited Harvard's system as a model for private institutions to follow. Since the '70s, the number of minority admits has steadily risen. Tests have become secondary, in part because educational experts concluded that they were far less objective than once thought; they are often biased in favor of the wealthy.

Performance on standardized tests, for example, is tightly linked with the quality of a test-taker's previous schooling. And while SAT II subject tests are the best predictors of college grades, Fitzsimmons says they are even more dependent on the quality of schooling than the SAT I.

So admissions officers try to use other test scores to gauge academic accomplishment, such as Advanced Placement (AP) tests. But Fitzsimmons acknowledges that even AP tests and other tests are all correlated with a person's socioeconomic status. Poorer schools may not have the resources to offer AP classes.

As to whether legacies and minority applicants receive undue special consideration, Fitzsimmons says that being the child of a parent who went to the College is "only a feather on a scale"--a factor taken into consideration at the end of the process if all other things are equal.

"We can have people from the very richest to the poorest backgrounds, but we look for people who have been able to use their opportunities to create a sense of accomplishment," he says.

Race, gender and socioeconomic status are all variables in the equation.