"Randomization" turned three this year, fulfilling the goal of former Dean of the College L. Fred Jewett '57 to diversify the student populations of Harvard's 12 residential Houses.

But with the decision to disallow student choice in housing firmly entrenched in College policy and student mindset, it was a milestone this year that no one seemed to notice.

Indeed, the first year of full randomization in the Houses elicited little more than a sigh from a student body that had largely decried the decision in angry protests just three years earlier.

But although many students and some House Masters remained quietly hopeful that randomization would be reviewed and perhaps even reversed, the likelihood at this point of any formal changes to the House system is slim, according to Dean of the College Harry R. Lewis '68.

In an interview with The Crimson last month, Lewis reiterated comments he had made at an April meeting of the Committee on House Life (COHL) that the policy would remain substantively unchanged through the duration of his tenure, adding only that the policy would likely be "tweaked."

That assertion, while not unexpected, came in the light of new data released by the College in April that showed a decline in the number of large blocking groups--those with 8 or more members--in which more than 50 percent of the students are in a particular ethnic group. And while the data also showed a surge in the number of large groups with a majority of varsity athletes, administrators reacted positively to the post-randomization demographics.

In particular, many officials saw the data as initial evidence that randomization had not had the adverse effects some critics forecasted. Last spring, a group of 26 House tutors, all minorities, sent an open letter to University officials saying that randomization had dissolved the support networks that certain prerandomized House communities had offered minority students.

Since the data indicated a trend showing blocking groups becoming increasingly diverse, Masters and administrators responded to the news with guarded optimism.

But critics--such as some House tutors--still raised concerns that the criteria used for the study might actually mask underlying problems. Because the College's figures measure only large blocking groups, they offer little indication of whether students are forming smaller groups that are ethnically homogeneous.

Eliot House Resident Scholar Nicky Sheats, a long-time resident tutor who helped write last year's letter, said his informal discussions with black students living in Eliot House suggest this is the case, and he hopes the College will continue to evaluate the effects of randomization policy.

But despite the downward trend in ethnic group representation in the groups, additional figures released by the College showed a strong upward trend in the size of student blocking groups between 1996 and 1999, strengthening calls to reduce the size of the blocks.

House Masters--seldom unified on particular issues--have presented a surprisingly solid front concerning the negative effects of large blocking groups. The worry, many have said, is that it becomes increasingly difficult to encourage residents to participate in House activities as the group sizes rise because students tend to have less need for the social opportunities offered by their House.

As a result, the data--which shows a notable increase in the number of blocking groups with the maximum number of 16 students--has given momentum to a proposal to reduce the maximum size of the groups.

Lewis, who oversees the House system as dean of the College, has pledged to give blocking size more formal consideration over the summer. He hopes to have the decision made by early next fall to allow first years as much time as possible to form appropriate-sized blocking groups.

Although Lewis refused to indicate his intentions, he said smaller blocking groups would afford the College increased flexibility in creating gender balance in the Houses. Reduced blocking group size would allow the College to lower the acceptable range of gender ratios that govern housing assignments.