Controlled School Choice with Hard Bounds: Existence of Fair and Non-wasteful Assignments Kentaro Tomoeda Harvard University
December 2012
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Main Results
I
Consider the model of controlled school choice employing type specific constraints with lower bounds and upper bounds, in which the set of feasible and fair assignments may be empty.
I
This research gives a sufficient condition on a schools’ priority profile for the existence of feasible assignments that are fair and non-wasteful.
I
Also introduces an algorithm which finds such an assignment under that sufficient condition.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Controlled School Choice
I
Controlled school choice programs are specific school choice programs that try to balance the racial or socioeconomic status.
I
Some of them introduce lower bounds and upper bounds of type specific quotas.
I
Examples include New York City high school match, Miami-Dade County Public Schools, and Chicago Public Schools.
I
With lower bounds and upper bounds, the set of feasible and fair assignments may be empty. ⇒ Need to (1) weaken the solution concept or (2) restrict priority structures.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Model A controlled school choice problem is (S, C , (qc )c∈C , PS , ÂC , T , τ, (q T , qT c )c∈C ) with c 1. a finite set of students S = {s1 , ..., sn }; 2. a finite set of schools C = {c1 , ..., cm }; 3. a capacity vector q = (qc1 , ..., qcm ), where qc is the capacity of school c ∈ C ; 4. a students’ preference profile PS = (Ps1 , ..., Psn ), where Ps is the strict preference relation of student s ∈ S over C ; 5. a schools’ priority profile ÂC = (Âc1 , ..., Âcm ), where Âc is the strict priority ranking of school c ∈ C over S; 6. a type space T = {t1 , ..., tk }; 7. a type function τ : S → T , where τ (s) is the type of student s; 8. for each school c, two vectors of type specific constraints t1 , ..., q tk ) such that = (q tc1 , ..., q tck ) and q T qT c c = (q c ∑ c ∑ q tc ≤ q tc ≤ qc for all t ∈ T , and t∈T q tc ≤ qc ≤ t∈T q tc . Kentaro Tomoeda
Controlled School Choice with Hard Bounds
An Assignment
An assignment µ is a function from the set C ∪ S to the set of all subsets of C ∪ S such that 1. µ(s) ∈ C for every student s; 2. |µ(c)| ≤ qc and µ(c) ⊆ S for every school c; 3. µ(s) = c if and only if s ∈ µ(c). µ is feasible if if for every school c, µ(c) respects constraints at c, i.e., |µ(c)| ≤ qc holds and q tc ≤ |µt (c)| ≤ q tc holds for every type t ∈ T and every student is assigned to a school.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Fairness and Non-wastefulness Definition 1. s justifiably claims an empty slot at c under µ if 1. cPs µ(s) and |µ(c)| < qc , τ (s)
2. q µ(s) < |µτ (s) (µ(s))|, and τ (s)
3. |µτ (s) (c)| < q c
.
µ is non-wasteful if there is no student who justifiably claims an empty slot at any school. Definition 2. s justifiably envies s 0 at c under µ if there exists another µ0 such that 1. µ(s 0 ) = c, cPs µ(s) and s Âc s 0 , 2. µ0 (s) = c, µ0 (s 0 ) 6= c, and µ0 (ˆs ) = µ(ˆs ) for all ˆs ∈ S \ {s, s 0 }. µ is fair across types (or fair) if there is no student who justifiably envies any other student. Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Non-existence of Feasible and Fair Assignments [Theorem 1] Theorem 1. (Abdulkadiro˘ glu and Ehlers (2008) and Ehlers, Hafalir, Yenmez and Yildirim (2011)) The set of feasible assignments that are fair across types may be empty in a controlled school choice problem.
capacities ceiling for t1 floor for t1 ceiling for t2 floor for t2
Âc1 s2 s1 s3 1 1 1 1 0
Âc2 s2 s1 s3 1 1 0 1 0
Kentaro Tomoeda
Âc3 s1 s2 s3 1 1 0 1 0
Ps1 c2 c3 c1
Ps2 c3 c2 c1
Ps3 c2 c3 c1
Controlled School Choice with Hard Bounds
Non-existence of Feasible and Fair Assignments [Theorem 1]
capacities ceiling for t1 floor for t1 ceiling for t2 floor for t2
Âc1 s2 s1 s3 1 1 1 1 0
Âc2 s2 s1 s3 1 1 0 1 0
Âc3 s1 s2 s3 1 1 0 1 0
Ps1 c2 c3 c1
Four feasible assignments: ( ) ( c1 c2 c3 c1 µ1 = , µ2 = s1 s2 s3 s1 ( ) ( c1 c2 c3 c1 µ3 = , µ4 = s2 s3 s1 s2
P s2 c3 c2 c1
c2 c3 s3 s 2 c2 c3 s1 s 3
P s3 c2 c3 c1
) , ) .
In each assignment, there is a student who justifiably envies another student. Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Solutions in the Literature
Two papers to analyze weaker solution concepts. I
Abdulkadiro˘glu and Ehlers (2008) showed that the set of feasible assignments that are fair for the same type and constrained non-wasteful is not empty.
I
Ehlers, Hafalir, Yenmez and Yildirim (2011) give another solution; they interpret the constraints as soft bounds and find fair and non-wasteful assignments.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
A Sufficient Condition: Common Priority Order for Every Type
In this paper, I don’t weaken the solution concept but give a sufficient condition for the existence. Definition 4. A schools’ priority profile ÂC has a common priority order for type t ∈ T if s Âc s 0 ⇔ s Âc 0 s 0 for any c, c 0 ∈ C and s, s 0 such that τ (s) = τ (s 0 ) = t. This is satisfied if priority structures are more coarse than the type classification and the single tie-breaking is used.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Main Theorem
Theorem 2. Suppose that the schools’ priority profile ÂC has a common priority order for every type t ∈ T . Then there exists a feasible assignment that is fair across types and non-wasteful.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
An Algorithm: CDAAR Under the sufficient condition, Controlled Student Proposing Deferred Acceptance Algorithm with Reproposal (CDAAR) achieves such an assignment. [Start] Construct an ordering f such that f (s) < f (s 0 ) ⇔ s Âc s 0 and students make proposals to schools in this order. Start from the empty assignment. 1. Student s with f (s) = 1 proposes to the most preferable school and s is assigned to that school if there is a feasible assignment. If rejected, s is in the rejected status at c. k. Student s with minimum f −1 (s) who is unassigned or whose assigned school is not the most preferable one among the schools with non-rejected status makes a proposal. If s justifiably claims an empty slot or envies someone, s is admitted, and if there was no empty slot, a student with the lowest priority is rejected. Otherwise, s is rejected. Kentaro Tomoeda
Controlled School Choice with Hard Bounds
An Algorithm: CDAAR
k. (contd) ˆs ’s status at cˆ changes from the rejected status to the non-rejected status if ˆs was in the rejected status and one of the following conditions holds (reproposal conditions): (i) ˆs ∈ Sτ (s) , f (ˆs ) > f (s) and ˆs was unfree at step k-1 but became free at step k. (ii) ˆs ∈ Sτ (s) , f (ˆs ) < f (s) and cˆ = νk (s). (iii) There is an empty slot at cˆ and ˆs has the highest priority for cˆ among students who justifiably claim an empty slot. [End] The algorithm ends when there is no student to make a proposal.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Sketch of Proof [Theorem 2]
I
First, show that the CDAAR algorithm ends in finite steps under the condition. I
I
Can show that there is no cycle of proposals.
Next, show that the achieved assignment is feasible, fair across types and non-wasteful. I
Can show that the enough reproposals are made.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Incentives
There is a negative result on the incentives of students. Theorem 3. Suppose that the schools’ priority profile ÂC has a common priority order for every type t ∈ T and that an ordering f of students in the CDAAR algorithm satisfies f (s) < f (s 0 ) ⇔ s Âc s 0 for some c ∈ C and for any s, s 0 such that τ (s) = τ (s 0 ). Then, there is an ordering f with which the CDAAR mechanism is not dominant strategy incentive compatible.
Kentaro Tomoeda
Controlled School Choice with Hard Bounds
Summary
I
Introduced a sufficient condition “common priority for the same type students” for the existence of feasible assignments that are fair and non-wasteful in the controlled school choice.
I
The proof is done by an algorithm CDAAR, which finds such an assignment under that sufficient condition.
I
CDAAR is not always incentive compatible with any possible ordering f .
Kentaro Tomoeda
Controlled School Choice with Hard Bounds