Report Says House Diversity Uneven

The non-ordered choice lottery system may have increased diversity in Harvard's upperclass houses, but campus-wide uniformity is still a long way away, according to a new Office of Housing study.

The study compares data on the changing student populations of several houses from fall 1989 to fall 1992.

Harvard administrators are using the study to illustrate the full effects of a switch away from the ordered choice system. The old system allowed first-years to submit their top three house choices ranked in order of preference, while the current system asks students to submit four housing choices in no particular order.

The report discloses percentages of student representation in houses in several areas, including athletic participation, socio-economic background, academic area and performance and minority composition. The report presents data from all undergraduate houses but does not identify particular houses.

The report shows differing degrees of improvement in all the major target areas.

"From the examination I've done of it, I think we've done fairly well in various ways," said L. Fred Jewett '57, dean of the College. "The first concern we had was about balance and diversity in the houses. I think the datasuggest movement has been made."

Jewett cited academics as an area he feels hasexperienced a noticeable shift.

For example, the highest percentage ofhumanities concentrators in any house has droppedfrom 47 percent in 1989 to 36 percent in 1992.Though this may be partly attributed to the dropin humanities concentrators overall--from 27percent to 21 percent of the class--variation fromthe overall average has also decreased.

In 1989, the variance between the houses withthe lowest and highest numbers of humanitiesconcentrators was 11 percent from the average. In1992, the lowest percentage of concentrators--15percent--varied from the average by only sixpoints.

The distribution of Group I students has alsobecome more equitable. In 1989, percentages ofGroup I scholars in houses ranged from 14 percentto two percent, with an average of seven percent.The 1992 range shows between 12 and five percentof overall house populations in Group I, while theaverage has risen to eight percent.

Jewett said he was pleased with the changes.

"I'm most concerned with the ones that are notthe most visible--academic areas, areas ofconcentration--and what those things represent inthe houses. The fundamental thing you hope will goon in a college is education," he said.

Jewett said he thought the most significantgains were in diversity of socio-economicbackground in the houses.

The distribution of students on scholarship hasbecome more equitable. In 1989, 42 percent ofstudents in houses were on scholarship, but onehouse had only 27 percent scholarship students,while another housed 48 percent.

In 1992, the lowest concentration ofscholarship students had risen to 35 percent andthe highest to 56 percent. Since the overall classaverage had risen only one point to 43 percent,the figures indicate a more uniform distributionof scholarship students.

But not all areas have improved as rapidly.

"The place that has had less movement has beenthe area of ethnic balance," Jewett said.

In fact, the distribution of minorities seemsto have become less equitable, even as thepercentage of minority students has increased.

In 1989, with 27 percent minority studentsoverall in the classes, minorities ranged from 18to 33 percent in various houses.

In 1992, with 33 percent overall minoritystudents, the range of distribution has widenedfrom 20 to 54 percent among the houses.

The range of distribution inside major ethnicgroups also reflects this trend.

For example, the highest inhouse concentrationof Hispanic students in 1992 ranges from 12percent to four percent. The eight-pointdifference is greater than in 1989, whendistributions ranged from seven to two percent.

The range of distribution of Asian students hasalso grown. In 1989, the percentages of Asianstudents in houses ranged from 5 to 19 percent; in1992, the percentages ranged from 10 to 30percent.

However, in the case of Hispanic and Asianstudents, the overall number of students hasrisen. This is not the case with Black students.

While the class percentage of Hispanic studentsrose from 5 to 6 percent and the percentage ofAsian students rose from 8 to 12 percent, theoverall percentage of Black students has remainedstatic at 8 percent.

And the range of distribution of Black studentsin the houses has widened as well, going from fourto 12 percent in 1989 to four to 17 percent in1992.

"I don't know why this is so," Jewett said,adding, "There's no question that non-orderedchoice doesn't cause it."

Associate Dean of the College Thomas A. Dingman'67 said that while the current system hasimproved diversity, Harvard is far from havingnon-differentiated house communities.

Jewett agreed, saying, "My own feeling is thateach house should be a microcosm of the Harvardcommunity. Non-ordered choice is going to produceless skewed results, but it is never going toproduce houses that are relatively similar inmakeup."

He said he doubts the housing process willchange in time for this spring's lottery.

"I feel that the current breakdown, althoughnot perfect, is not a failure. We moved in theright direction," he said

Jewett cited academics as an area he feels hasexperienced a noticeable shift.

For example, the highest percentage ofhumanities concentrators in any house has droppedfrom 47 percent in 1989 to 36 percent in 1992.Though this may be partly attributed to the dropin humanities concentrators overall--from 27percent to 21 percent of the class--variation fromthe overall average has also decreased.

In 1989, the variance between the houses withthe lowest and highest numbers of humanitiesconcentrators was 11 percent from the average. In1992, the lowest percentage of concentrators--15percent--varied from the average by only sixpoints.

The distribution of Group I students has alsobecome more equitable. In 1989, percentages ofGroup I scholars in houses ranged from 14 percentto two percent, with an average of seven percent.The 1992 range shows between 12 and five percentof overall house populations in Group I, while theaverage has risen to eight percent.

Jewett said he was pleased with the changes.

"I'm most concerned with the ones that are notthe most visible--academic areas, areas ofconcentration--and what those things represent inthe houses. The fundamental thing you hope will goon in a college is education," he said.

Jewett said he thought the most significantgains were in diversity of socio-economicbackground in the houses.

The distribution of students on scholarship hasbecome more equitable. In 1989, 42 percent ofstudents in houses were on scholarship, but onehouse had only 27 percent scholarship students,while another housed 48 percent.

In 1992, the lowest concentration ofscholarship students had risen to 35 percent andthe highest to 56 percent. Since the overall classaverage had risen only one point to 43 percent,the figures indicate a more uniform distributionof scholarship students.

But not all areas have improved as rapidly.

"The place that has had less movement has beenthe area of ethnic balance," Jewett said.

In fact, the distribution of minorities seemsto have become less equitable, even as thepercentage of minority students has increased.

In 1989, with 27 percent minority studentsoverall in the classes, minorities ranged from 18to 33 percent in various houses.

In 1992, with 33 percent overall minoritystudents, the range of distribution has widenedfrom 20 to 54 percent among the houses.

The range of distribution inside major ethnicgroups also reflects this trend.

For example, the highest inhouse concentrationof Hispanic students in 1992 ranges from 12percent to four percent. The eight-pointdifference is greater than in 1989, whendistributions ranged from seven to two percent.

The range of distribution of Asian students hasalso grown. In 1989, the percentages of Asianstudents in houses ranged from 5 to 19 percent; in1992, the percentages ranged from 10 to 30percent.

However, in the case of Hispanic and Asianstudents, the overall number of students hasrisen. This is not the case with Black students.

While the class percentage of Hispanic studentsrose from 5 to 6 percent and the percentage ofAsian students rose from 8 to 12 percent, theoverall percentage of Black students has remainedstatic at 8 percent.

And the range of distribution of Black studentsin the houses has widened as well, going from fourto 12 percent in 1989 to four to 17 percent in1992.

"I don't know why this is so," Jewett said,adding, "There's no question that non-orderedchoice doesn't cause it."

Associate Dean of the College Thomas A. Dingman'67 said that while the current system hasimproved diversity, Harvard is far from havingnon-differentiated house communities.

Jewett agreed, saying, "My own feeling is thateach house should be a microcosm of the Harvardcommunity. Non-ordered choice is going to produceless skewed results, but it is never going toproduce houses that are relatively similar inmakeup."

He said he doubts the housing process willchange in time for this spring's lottery.

"I feel that the current breakdown, althoughnot perfect, is not a failure. We moved in theright direction," he said