Academia.eduAcademia.edu

Figure 3 – uploaded by Kaire Põder

See full PDFdownloadDownload figure
Table 5: Welfare of the students: Unassigned students We cannot include preference utilisation in the case of actual Tallinn mechanism into com- parison, because the ordered preferences are not reported. Thus our second-best empirical

Table 5 Welfare of the students: Unassigned students We cannot include preference utilisation in the case of actual Tallinn mechanism into com- parison, because the ordered preferences are not reported. Thus our second-best empirical

Related Figures (4)

Table 1: Properties of the mechanisms
Table 1: Properties of the mechanisms
Table 4: Welfare of the society: Priority utilisation Tallinn mechanism priorities students by how far they live from the school. Thus, technically we may say that political aim is to minimise the distance that students need to travel to school. In Table 4 comparison are shown on yearly bases. Overall we see that by applying an algorithmic approach we do not perform more that 10% worse than Tallinn. However, considering the number of unassigned students in Tallinn it is huge (Table 5), 16% tp 18% on average, when compared to any mechanism it is at least three times worse. A limitation to current result can be the diligence in updating the status of student’s assignment is poor and the matching data don’t reflect all matched (and accepted) pairs.
Table 4: Welfare of the society: Priority utilisation Tallinn mechanism priorities students by how far they live from the school. Thus, technically we may say that political aim is to minimise the distance that students need to travel to school. In Table 4 comparison are shown on yearly bases. Overall we see that by applying an algorithmic approach we do not perform more that 10% worse than Tallinn. However, considering the number of unassigned students in Tallinn it is huge (Table 5), 16% tp 18% on average, when compared to any mechanism it is at least three times worse. A limitation to current result can be the diligence in updating the status of student’s assignment is poor and the matching data don’t reflect all matched (and accepted) pairs.
Table 7: Welfare of the students: Siblings
Table 7: Welfare of the students: Siblings
Table 6: Welfare of the students: Received first preference In general the number of students who received their first preference is susceptible to manip- ulation. As we saw from the example students might not want to report their true preferences. While this is certainly true we are unsure how many students or families actually are able to compute the best strategy. In general, as literature shows the truthful preference revelation has merits on his own, i.e. not forcing parents to find the strategy for the optimal manipulation. In aggregate (Table 6) we see that with ScRSD mechanism we would find a suboptimal allocation, thus suggesting that there is room for manipulation.
Table 6: Welfare of the students: Received first preference In general the number of students who received their first preference is susceptible to manip- ulation. As we saw from the example students might not want to report their true preferences. While this is certainly true we are unsure how many students or families actually are able to compute the best strategy. In general, as literature shows the truthful preference revelation has merits on his own, i.e. not forcing parents to find the strategy for the optimal manipulation. In aggregate (Table 6) we see that with ScRSD mechanism we would find a suboptimal allocation, thus suggesting that there is room for manipulation.
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved