Over the course of the fall semester, the Transition Subcommittee of the CUL discussed various possibilities for moving from the current free agent system to the Williams House System.  Our proposal addresses several aspects of this transition.  First, it recommends a process through which students will choose both their neighborhood and room.  Second, it recommends a process through which faculty and staff will join the House System.  After describing these processes, the proposal explains the reasoning that led the committee to recommend them.  We hope that our proposal achieves the goals laid out in last yearÕs CUL report, including the preservation of the most valuable attributes of the current system.

 

I.  Recommended Transition

 

      A.  Student Selection of Neighborhoods and Rooms

 

The goal of the transition, a stable and viable House System, is designed to benefit all students; in attempting to achieve this goal, this subcommittee has attempted to design a transition that mixes the best features of free agency and neighborhood housing.  Under the proposed model there are two separate draws, one for neighborhoods and one for rooms.  Students who are not able to attend either or both of these draws may have a proxy choose in their place, ensuring that all students are able to participate in the transition.  We propose that the draws be conducted in the following manner:

 

(1)     Housing booklets will go out to all Williams students on February 10th, 2006. All current Williams students will be entered into a neighborhood draw including those students planning to study abroad and those who have applied to be JAs.  Students will form groups of up to six (6) students.  Groups can be comprised of students from any class year and from any dorm.  These booklets will be due back to the Office of Campus Life by February 24th. 

 

(2)     Over the next week, the Office of Campus Life will use the existing algorithm to generate pick numbers for the neighborhood draw.  These numbers will be available to students by March 2nd.  Rising seniors will have higher pick numbers than rising juniors, who will in turn have higher pick numbers than rising sophomores.  Groups comprised of mixed classes will have their years averaged and will be placed appropriately; for instance, a group consisting of a junior and a senior would be assigned to the bottom of the senior class, but above a group consisting of two juniors and a senior.

 

(3)     The neighborhood draw will be held at the Log between March 7th and March 9th, and will be nearly identical to past free-agent draws except for the fact that students will be selecting a region of the campus rather than individual rooms.  Neighborhoods will be capped by gender and by class year. Spaces will also be set aside in each neighborhood for students who are picking in as Junior Advisor and Fall Term Abroad applicants. Thus individuals choosing to participate in these programs that affect housing can still become a part of the House System with their friends and peers, and the transition to the House System will not affect the timeline of these pre-existing programs.

 

(4)     Between April 10th and April 12th, each neighborhood will hold a second draw in which its members will choose specific rooms.  These room draws will again give precedence to seniority, but within seniority cohorts the order in which groups pick could be determined in either of the following ways:

 

a)     Campus Life will generate a new set of random numbers to determine the order in which groups will choose their rooms within the neighborhood.

 

b)    The order in which groups choose their rooms will be determined by reversing the order in which the groups entered the neighborhood.  Thus groups that enjoy an early pick in the neighborhood draw will choose late in the room draw, whereas groups that choose late in the neighborhood draw will enjoy an early choice of rooms.

 

The Transition Subcommittee has determined that options 4a and 4b would work equally well, and the prevailing opinion of both the Transition Subcommittee and the online student poll is equally divided between these methods.  There is a certain appeal among some students for assigning pick numbers randomly in both the neighborhood and the room draws (4a), since this method most closely resembles the old lottery number generation under the all-campus room draw.  However, it also has the added stress of forcing students to go through the lottery number process twice, each time with uncertain results, and will result in some groups receiving poor picks for both their neighborhood and rooms.  The algorithm that assigns pick numbers in the room draw by reversing the pick order from the neighborhood draw (4b) is potentially more equitable, but may be perceived as a form of engineering on the part of the CUL.  The full CUL could not achieve unanimity regarding which of these two options is preferable.  The majority favors reverse-ordering the draws (4b), on the grounds that no student would lose in both lotteries, and that after the initial assignment of numbers all parties would know where they stand and would be able to plan their choices without the additional anxiety of having to await an additional round of random number generation.  The differences between the two options are treated in greater detail below in the section entitled ÒProcess and Background,Ó which we hope the Dean may find useful in making a decision regarding this aspect of the transition.

 

B.  Faculty and Staff Affiliation

 

Participation of faculty will be important to the success of the new House System. The Transition Subcommittee therefore proposes that all faculty members be assigned a neighborhood affiliation.  Faculty from each department should be distributed roughly equally across the five neighborhoods, so that each neighborhood has a representative cross-section of the full faculty.  For those departments that have fewer than five members, each member should be in a different neighborhood.  Discussion with current faculty members on the CUL indicates that most faculty will have little opinion or knowledge of the neighborhoods prior to the start of the House System, and will not object to being affiliated in this manner. Attempts should be made, however, to accommodate individual faculty members who do express a strong preference for being affiliated with a particular neighborhood.

 

All staff should be welcomed and actively invited, but not required, to receive a neighborhood affiliation.  Individual staff members should be encouraged to consider carefully the extent to which they would like to participate in the House System.  Some hourly staff, for example, might enjoy being affiliated with the neighborhood in which they work.  Others, however, might feel that affiliation would be an additional responsibility that they should not be asked to bear without compensation.  Participation of salaried administrators who provide services to the entire campus should also be voluntary.  All administrators who want to be affiliated will be randomly assigned to a neighborhood, unless they bring a specific preference to the attention of the CUL.

 

We propose that the faculty and staff affiliation process be completed by the end of the fall semester (December 2005).  Having a list of faculty/staff neighborhood affiliations may be a factor that will help some students decide upon their own neighborhood.

 

II.  Process and Background

 

The Transition Subcommittee met between September and November 2005 to discuss the transition from the current free-agent system to the Williams College House System. The Committee was aware that our recommendation is likely to be intensely scrutinized by the student body, partially because it affects housing, but also because it is the first major step toward the implementation of the House System, which up until this point has been largely abstract. We also realized there is likely to be anxiety and opposition due to the fact that we are changing the existing institution of the All-Campus Room Draw, something which is comforting to students in its familiarity. With these things in mind, the committee strove to find a transition process that would be fair and equitable to everyone, and would provide students with as much control within the process as we could reasonably allow and still have it function smoothly. Another of our initial goals was to create as simple a process as possible, but to a certain extent we also felt that a somewhat more complex system would be required to meet our other goals of fairness, equity, and increased student control. We recognize that no transition process could be perfect, but believe that our recommendation provides the basis for a successful implementation of the Williams House System.

 

In the remaining sections of this document we describe the evolution of the thinking that led the Transition Subcommittee to its recommendation regarding the best way to allow students to choose their neighborhoods and rooms.

 

A.  Initial Considerations

 

Special emphasis was placed on making the transition as smooth as possible, minimizing uncertainty and frustration.  We first passed over systems at the extremes, a purely random assignment and a pure room draw, as being against the values of the old and new system, respectively.  We then rejected a variety of hybrid schemes, in which neighborhood populations would be assigned based upon the gathered preferences of individual students, because they were overly complex and would rely too heavily on the discretion of the Office of Campus Life.  Our final design, a two-lottery system with limited constraints, is one that we feel is preferable because it maximizes simplicity, transparency, and student choice.

 

B.  Random Assignment

 

A purely random assignment of student groups to neighborhoods would ensure social diversity, but it would not be able to take into account any student preferences whatsoever.  A random system would also be unfair to current students who were admitted when the free agent system was in place and were unaware of the impending change to neighborhood housing.  Furthermore, a purely random system would be unpopular with the Williams student body and unlikely to encourage upperclassmen to participate fully in the new House System.  In light of these objections, and several others, the Transition Subcommittee rejected the possibility of random assignment and aimed to preserve as much student choice as possible.

 

C.  Traditional Room Draw

 

While turning away from a purely random-assignment model, the subcommittee realized that it needed to make several changes to the current room draw procedure in order to make neighborhood creation feasible.  Though it would be possible to have one last, all-campus room draw in which students chose both one-year rooms and multi-year neighborhoods simultaneously, the subcommittee felt that such a system might not be fair.  Such a scheme would give an undue benefit to the individual who was pick number one while unduly penalizing other individuals with less advantageous picks.  For instance, the sophomore with the last pick would not only receive an unfavorable room, but might also wind up in a neighborhood that he or she considers undesirable.  Additionally, the current room draw model might fail to produce an outcome that balances class-year distribution across neighborhoods, which is a requirement for the successful launch of the house system.  Based on the expectation of current students that sophomores live in Mission Park and juniors live in Greylock, an unbridled room draw would likely result in some neighborhoods having a disproportionate number of members from one class.  The subcommittee therefore decided to consider ways to address these issues without overly restricting studentsÕ ability to choose their rooms.

 

D.  Neighborhood Assignment by The Office of Campus Life

 

Having rejecting the two extremes for the reasons outlined above, the subcommittee began to examine hybrid systems.  The first system considered was one under which pick groups would rank neighborhoods on a room draw form and the Office of Campus Life would attempt to guarantee that students were not placed in their least-favored neighborhood.  Under this plan, each neighborhood would then conduct its own draw for individual rooms.  This method had the advantage of satisfying more students than a purely random method of cluster assignment.  It also preserved a certain degree of randomization, as the Office of Campus Life would not be able to place all students in their first-choice neighborhood.  However, the subcommittee had serious reservations about this algorithm.  The system was unnecessarily opaque, as members of the Office of Campus Life, rather than the students themselves, would be ultimately responsible for neighborhood assignment.  Thus although a pick group might be willing to select into a less favorable neighborhood in order to live near their friends, the Office of Campus Life would not be able to satisfy this request due to the additional complexity that would result.  Most importantly, this system did not have any protections for students who were assigned to neighborhoods that they consider undesirable.  Under this system, it would be possible for groups to be assigned to one of their least favorite neighborhoods and then to receive an unfavorable pick in the lottery for individual rooms.

 

The subcommittee next considered a system that would randomly assign students to neighborhoods, but would then attempt to reward those students who were not assigned to their first choice.  The Office of Campus Life would attempt to match students with their choice of neighborhood, and if the system were unable to satisfy student requests then it would give prime rooms within the neighborhood to students who did not particularly wish to be assigned there.   Despite this improvement, many of the above problems remained.  The system was at least as opaque, and perhaps even more so, since the assignment of individual rooms would have to be done in Hopkins Hall.  Additionally, this change still did not address the fact that two pick groups might prefer to be in a neighborhood they considered less desirable if it meant that they could be together.

 

Over the course of several meetings, then, the subcommittee decided to abandon the idea that students would be assigned to neighborhoods by the Office of Campus Life.  While such a procedure might achieve some of the short-term goals of the neighborhood system by integrating different types of student groups, it would be widely unpopular on campus, and would create excessive logistical problems.

 

E.   Two-Draw Algorithm

 

In light of the above considerations, the subcommittee examined the possibility of having two separate draws, one for neighborhoods and one for rooms.  This plan would allow students to choose their own neighborhoods according to a draw formula similar to the one with which they are already familiar, and then to select individual rooms.  The subcommittee felt that this plan would achieve the best balance between the two goals of giving students a significant degree of choice and ensuring that the initial neighborhood populations reflect the diversity of the student body.  Because the pick numbers for the neighborhood draw would be randomly generated, there would be no guarantee that different pick groups would end up in the same neighborhood and thus there would be a sufficient degree of population mixing.  At the same time, however, the plan would allow students to select their neighborhood for themselves in light of their own assessment of their housing priorities.

 

After the subcommittee decided upon recommending two sequential draws, it identified and discussed two different ways in which these draws could be conducted.

 

Idea A:  Unrelated Draws

 

One possibility is to hold two completely distinct and unrelated draws.  If students feel that the neighborhood affiliation and room choice should be treated as two independent events, then two independent sets of random numbers could be generated.  Using this algorithm, the Office of Campus Life would generate one set of random numbers for the neighborhood draw, and then generate a completely new set of random numbers for the room draw within each neighborhood.  Such a system would have the advantage of preserving the chance for an individual to receive both a very high neighborhood draw number and a very high room draw number, an aspect that might be pleasing to risk-seeking students.  Also, the familiarity of draws conducted on the basis of randomly assigned pick numbers might help make this yearÕs draws more successful in trying to please as many students as possible.  This system would also enable students to account for the actions of their friends when selecting neighborhoods and rooms.  This algorithm would combine elements of choice and random selection, be highly transparent, and could be easily implemented as well.

 

Idea B: Inversely Related Draws

 

A second possibility is to relate the two draws inversely, so that the order in which students choose rooms would be the inverse of the order in which they choose into the neighborhood.  This would have the advantage of ensuring that a student with an undesirable neighborhood pick would be compensated with a more desirable pick in the room draw.  Under this system, every student would be guaranteed a favorable room or a favorable neighborhood, an aspect that would be pleasing to risk-averse students.  This system would also minimize student uncertainty and anxiety, since pick groups would know at the moment of neighborhood selection roughly where they would fall in the subsequent room draw, and would therefore have an intuition as to what sort of rooms would be available to them within each neighborhood.  Like the unrelated draws option, the inversely related draws option would be a transparent process that would enable students to account for the actions of their friends, and could also be easily implemented by the Office of Campus Life.

 

 

      F.  Rising Seniors

 

The Transition Subcommittee also wishes to address various proposals that would give rising seniors special treatment in the upcoming transition to the Williams House System. Some rising seniors feel that the college and CUL ÒoweÓ them the opportunity to use the same room draw procedure that they have used in previous years. There is some concern that if these wants are not catered to in the initial transition then there will be little support from the class of 2007 for the House System next year.

 

Discussions in the Transition Subcommittee led to the conclusion that it would not be in the best interest of the college to approve a transition procedure that would favor one class year more than any other. Furthermore, it was felt that putting undue emphasis (in the form of special treatment in the transition process) on the senior class might create resentment among the other class years, and would not necessarily increase the appeal of the Williams House System.  While the senior class will certainly be important to neighborhood life, there are several reasons why we feel that designing an algorithm around pleasing the senior class would not be wise: