Researchers have found considerable evidence that information about school performance affects people's choices about which schools to send their children to and even where to live. In contrast, little attention has been paid to the effects of school performance information on people's political behavior. Yet Hirschman (1970) famously highlighted the importance of taking seriously not only economic forces, but also the role of "political mechanisms", that is, "non-market forces" or "voice", in analyzing people's responses to school performance and the implications of these responses for school outcomes. This dissertation explores the effect of information about student and school performance on people's political attitudes and behavior. I first present findings from an original dataset of school board elections in Florida that indicate that voters fail to punish school board incumbents in response to information signaling poor school performance. There is even evidence that voters sometimes reward incumbents for failure. I next analyze a dataset that links student test scores in England to a subsequent survey, and find that that informational signals about individual student performance can have long−lasting effects on parental behavior. Finally, I analyze the results of a survey experiment administered to a nationally representative sample of Americans, and find that information about the relative performance of local schools depresses average perceptions of local school quality and increases support for school reforms. / Government

Στο πλαίσιο της παρούσας διδακτορικής διατριβής μελετώνται υπολογιστικά ζητήματα που προκύπτουν από τη θεωρία της κοινωνικής επιλογής. Ένα από τα κύρια θέματα της θεωρίας αυτής είναι οι εκλογές. Τα προβλήματα που σχετίζονται με τις εκλογές ανήκουν στη θεωρία ψηφοφοριών όπου βασικό πρόβλημα είναι η εύρεση του νικητή των εκλογών όταν έχουμε ως δεδομένες τις προτιμήσεις των ψηφοφόρων. Στη βιβλιογραφία υπάρχουν αρκετοί κανόνες ψηφοφορίας βάσει των οποίων γίνεται ο υπολογισμός της κατάταξης μιας ψηφοφορίας και της ανάδειξης του νικητή. Η θεωρία των ψηφοφοριών αποτελεί ένα σημαντικό κλάδο της θεωρίας της κοινωνικής επιλογής με άμεσες εφαρμογές στην κοινωνία, καθώς οι κανόνες αυτοί εφαρμόζονται στην πράξη σε βουλευτικές, δημοτικές ή τοπικές εκλογές, καθώς επίσης και σε αρκετές επιτροπές σχετικές με την λήψη συλλογικών αποφάσεων. Στη συγκεκριμένη διατριβή μελετούμε πρώτα τους κανόνες ψηφοφορίας που πρότειναν ο Dodgson και ο Young, οι οποίοι είναι σχεδιασμένοι έτσι ώστε να εντοπίζουν τον υποψήφιο που είναι διαισθητικά πιο κοντά στο να είναι ο νικητής σύμφωνα με το κριτήριο του Condorcet. Σύμφωνα με το κριτήριο αυτό, νικητής θα πρέπει να είναι ο υποψήφιος που έχει την προτίμηση της πλειοψηφίας έναντι των άλλων υποψήφιων. Ωστόσο κάτι τέτοιο δεν μπορεί να υπολογιστεί πάντα, καθώς οι προτιμήσεις της πλειοψηφίας μπορεί να είναι κυκλικές. Για παράδειγμα σε μια εκλογή με 3 υποψηφίους, ο υποψήφιος a να προτιμάται σε σχέση με τον b, ο b να προτιμάται σε σχέση με τον c, αλλά ο c να προτιμάται σε σχέση με τον a. Στην περίπτωση αυτή δεν μπορεί να καθοριστεί ο νικητής των εκλογών σύμφωνα με το κριτήριο του Condorcet. Για το λόγο αυτό χρησιμοποιούμε τους κανόνες ψηφοφορίας που προτάθηκαν από τους Dodgson και Young, οι οποίοι παρέχουν μια διαφορετική έννοια εγγύτητας ενός υποψήφιου στο να αναδειχθεί νικητής κατά Condorcet. Οι συγκεκριμένοι κανόνες αποτελούν ένα σημαντικό κομμάτι της βιβλιογραφίας της θεωρίας της Κοινωνικής Επιλογής διότι διαισθητικά παρουσιάζουν υψηλή εγγύτητα με τον κανόνα του Condorcet. Πιο συγκεκριμένα και σύμφωνα με τον κανόνα του Dodgson ισχύουν τα εξής: δεδομένου ενός συνόλου προτιμήσεων των ψηφοφόρων, ο βαθμός ενός υποψηφίου ορίζεται ως ο ελάχιστος αριθμός των ανταλλαγών που πρέπει να γίνουν μεταξύ γειτονικών υποψηφίων στην κατάταξη των ψηφοφόρων έτσι ώστε ο συγκεκριμένος υποψήφιος να αναδειχθεί νικητής κατά Condorcet. Ο υποψήφιος που έχει τον ελάχιστο βαθμό Dodgson είναι νικητής κατά Dodgson. Αντίστοιχα, σύμφωνα με τον κανόνα του Young ο βαθμός ενός υποψηφίου είναι το μέγεθος του μεγαλύτερου υποσυνόλου ψηφοφόρων έτσι ώστε, αν ληφθούν υπόψη μόνο αυτά τα ψηφοδέλτια, ο συγκεκριμένος υποψήφιος να γίνεται νικητής κατά Condorcet. Ο υποψήφιος με το μέγιστο βαθμό Young είναι νικητής κατά Young. Δυστυχώς, ο υπολογισμός του βαθμού ενός δεδομένου υποψηφίου είναι δύσκολος είτε με τον κανόνα του Dodgson είτε με τον κανόνα του Young. Για αυτό το λόγο τα υπολογιστικά ζητήματα που προκύπτουν και μελετώνται στην παρούσα διατριβή αφορούν στην προσέγγιση των δυο αυτών κανόνων ψηφοφορίας. Πιο συγκεκριμένα παρουσιάζονται δύο αλγόριθμοι που προσεγγίζουν το βαθμό ενός υποψηφίου σύμφωνα με τον κανόνα του Dodgson: ένας άπληστος αλγόριθμος και ένας αλγόριθμος βασισμένος σε γραμμικό πρόγραμμα. Οι δυο αυτοί αλγόριθμοι έχουν λόγο προσέγγισης H_{m-1}, όπου m είναι ο αριθμός των υποψηφίων και H_{m-1} είναι ο (m-1)-ος αρμονικός αριθμός. Επίσης, αποδεικνύεται ότι δεν υπάρχει αλγόριθμος πολυωνυμικού χρόνου που να προσεγγίζει το βαθμό Dodgson κατά λογαριθμικό παράγοντα, εκτός εάν υπάρχουν για τα προβλήματα της κλάσης NP αλγόριθμοι ψευδο-πολυωνυμικού χρόνου. Παρότι διαισθητικά υπερέχει ο άπληστος αλγόριθμος, στη συγκεκριμένη διατριβή υποστηρίζουμε ότι ο αλγόριθμος που είναι βασισμένος σε γραμμικό πρόγραμμα έχει πλεονέκτημα από την οπτική της θεωρίας της κοινωνικής επιλογής. Επιπλέον, αποδεικνύεται ότι ο υπολογισμός κάθε λογικής προσέγγισης της κατάταξης που παράγεται από τον κανόνα Dodgson είναι ένα υπολογιστικά δύσκολο πρόβλημα, γεγονός που εξηγεί, από την πλευρά της θεωρίας της πολυπλοκότητας, το ότι έχουν παρατηρηθεί μεγάλες διαφορές κατά τη σύγκριση εκλογών Dodgson με απλούστερους κανόνες ψηφοφορίας που εμπεριέχονται στη βιβλιογραφία της θεωρίας της κοινωνικής επιλογής. Τέλος, αποδεικνύεται ότι το πρόβλημα υπολογισμού του βαθμού ενός υποψηφίου σύμφωνα με τον κανόνα του Young είναι επίσης υπολογιστικά δύσκολο. Αυτό οδηγεί σε ένα αποτέλεσμα μη προσεγγισιμότητας για την κατάταξη υποψηφίων σε μια εκλογή σύμφωνα με τον κανόνα του Young. Αν και ο κανόνας του Dodgson είναι ένας από τους πιο καλά μελετημένους κανόνες ψηφοφορίας, παρουσιάζει σοβαρές ελλείψεις, τόσο από υπολογιστικής άποψης --- είναι υπολογιστικά δύσκολο ακόμη και να προσεγγιστεί ο βαθμός Dodgson ενός υποψηφίου --- όσο και από την πλευρά της κοινωνικής επιλογής, καθώς αποτυγχάνει σε βασικές επιθυμητές ιδιότητές της, όπως είναι η μονοτονία και η ομοιογένεια. Ωστόσο, αυτό δεν αποκλείει την ύπαρξη προσεγγιστικών αλγορίθμων για τον κανόνα του Dodgson που είναι μονότονοι ή ομοιογενείς, οπότε τίθεται το ερώτημα ύπαρξης τέτοιων αλγόριθμων. Στη διατριβή αυτή δίνονται οριστικές απαντήσεις στα ερωτήματα αυτά. Παρουσιάζεται ένας μονότονος αλγόριθμος εκθετικού χρόνου που πετυχαίνει λόγο προσέγγισης του βαθμού Dodgson ίσο με 2 και το αποτέλεσμα συμπληρώνεται με ένα αυστηρό αντίστοιχο κάτω φράγμα. Παρουσιάζεται επίσης ένας μονότονος προσεγγιστικός αλγόριθμος πολυωνυμικού χρόνου με λόγο O(logm) (όπου m είναι ο αριθμός των υποψηφίων): και στην περίπτωση αυτή το αποτέλεσμα είναι βέλτιστο, λόγω ύπαρξης αντίστοιχου κάτω φράγματος. Επιπλέον, αποδεικνύεται ότι μια μικρή παραλλαγή σε ένα γνωστό κανόνα ψηφοφορίας δίνει ένα μονότονο, ομοιογενή, O(m logm)-προσεγγιστικό αλγόριθμο πολυωνυμικού χρόνου, με τον καλύτερο δυνατό λόγο προσέγγισης, ακόμη και αν αυτό που μας ενδιαφέρει είναι μόνο η ομοιογένεια. Τέλος, μελετώνται διάφορες πρόσθετες ιδιότητες κοινωνικής επιλογής, για τις οποίες δεν υπάρχει αλγόριθμος με λόγο προσέγγισης που να εξαρτάται μόνο από το m . Αυτά τα αποτελέσματα της προσεγγισιμότητας του κανόνα αυτού αποτελούν σημαντική προσφορά της διατριβής καθώς μπορούν να θεωρηθούν ως κανόνες που υπολογίζονται σε πολυωνυμικό χρόνο και πληρούν μάλιστα επιθυμητές κοινωνικές ιδιότητες, ενώ ταυτόχρονα διέπονται και από την φιλοσοφία του κανόνα που θέσπισε ο Dodgson. Ένα άλλο σημαντικό υπολογιστικό πρόβλημα με το οποίο ασχολείται η παρούσα διατριβή είναι αυτό της δωροδοκίας των εκλογών. Στο πρόβλημα της δωροδοκίας μπορεί κάποιος με δεδομένο ένα χρηματικό προϋπολογισμό να αλλάξει τις προτιμήσεις των ψηφοφόρων ώστε να αναδειχθεί νικητής των εκλογών ο υποψήφιος της αρεσκείας του. Στη συγκεκριμένη διατριβή μελετώνται κανόνες ψηφοφορίας που βασίζονται στη βαθμολόγηση των υποψηφίων. Πιο συγκεκριμένα μελετάται η τάξη των κανόνων ψηφοφορίας όπου κάθε ψηφοφόρος εκχωρεί κ βαθμούς στον υποψήφιο που προτιμά ως πρώτο, λ βαθμούς στον υποψήφιο που προτιμά ως δεύτερο και 0 βαθμούς σε όλους τους υπόλοιπους υποψηφίους. Αποδεικνύεται ότι για αυτήν την τάξη των κανόνων βαθμολόγησης η δωροδοκία είναι ένα υπολογιστικά δύσκολο πρόβλημα. Στους κανόνες που βασίζονται στην βαθμολόγηση των υποψηφίων περιλαμβάνονται αυτοί της πλειοψηφίας και της 2-έγκρισης όπου μια βέλτιστη στρατηγική δωροδοκίας μπορεί εύκολα να υπολογιστεί, καθώς επίσης και ο κανόνας της 3-έγκρισης όπου η δωροδοκία είναι ένα υπολογιστικά δύσκολο πρόβλημα. Λαμβάνοντας υπόψιν την πολυπλοκότητα αυτών των κανόνων εξάγεται το συμπέρασμα ότι οι κανόνες που μελετήθηκαν είναι εκ των πιο απλών που δεν είναι ευάλωτοι στη δωροδοκία. / In this PhD thesis we study computational problems arising from the theory of social choice. One main aspect of Computational Social Choice is voting theory. The most important problem of voting theory is the computation of the winner of the elections when we have as input the preferences of the voters. In the literature there are many voting rules according to which the computation of the winner of the elections is done. Voting theory is a seminal subject in the Computational Social Choice theory with applications in the society. Voting rules are widely used in government and municipal or local elections and also in committees for taking decisions. In this thesis we start by considering voting rules proposed by Dodgson and Young. These rules are both designed to find an alternative closest to being a Condorcet winner. According to the Condorcet criterion, the winner of the elections should be the one that the majority of the voters prefer in relation to every other candidate. Unfortunately, the preferences of the majority may be circular. For example, in an election with 3 candidates, candidate a is preferred to b by the majority and b is preferred in relation to c, but c is preferred to a. Then a Condorcet winner does not exist. Each of these voting rules provide a different notion of proximity of how close they are to Condorcet rule. In the Dodgson rule the score of a candidate, given a set of preferences, is the minimum number of exchanges between adjacent candidates in order for the specific candidate to become a Condorcet winner. The Dodgson winner is the candidate with the minimum Dodgson score. In the Young rule the score of a candidate is the size of the largest subset of voters, when taking in account only these votes, the specific candidate becomes a Condorcet winner. The Young winner is the candidate with the maximum Young score. The score of a given alternative is known to be hard to compute under either rule and so the computational problems that arise and we consider are related to the approximation of the voting rules proposed by Dodgson and Young. We put forward two algorithms for approximating the Dodgson score: a combinatorial, greedy algorithm and an LP-based algorithm, both of which yield an approximation ratio of H_{m-1}, where m is the number of alternatives and H_{m-1} is the (m-1)st harmonic number. We also prove that there is no polynomial time algorithm that approximates the Dodgson score by (1/2-ε)lnm, unless problems in NP have quasi-polynomial time algorithms. Despite the intuitive appeal of the greedy algorithm, we argue that the LP-based algorithm has an advantage from a social choice point of view. Further, we demonstrate that computing any reasonable approximation of the ranking produced by Dodgson's rule is NP-hard. This result provides a complexity-theoretic explanation of sharp discrepancies that have been observed in the social choice theory literature when comparing Dodgson elections with simpler voting rules. Also, we show that the problem of calculating the Young score is NP-hard to approximate by any factor. This leads to an inapproximability result for the Young ranking. Although Dodgson's rule is one of the most well-studied voting rules, it suffers from serious deficiencies, both from the computational point of view --- it is NP-hard even to approximate the Dodgson score within sublogarithmic factors --- and from the social choice point of view --- it fails basic social choice desiderata such as monotonicity and homogeneity. However, this does not preclude the existence of approximation algorithms for Dodgson that are monotonic or homogeneous, and indeed it is natural to ask whether such algorithms exist. In this thesis we give definitive answers to these questions. We design a monotonic exponential-time algorithm that yields a 2-approximation to the Dodgson score, while matching this result with a tight lower bound. We also present a monotonic polynomial-time O(logm)-approximation algorithm (where m is the number of alternatives); this result is tight as well due to a complexity-theoretic lower bound. Furthermore, we show that a slight variation on a known voting rule yields a monotonic, homogeneous, polynomial-time O(m logm)-approximation algorithm, and establish that it is impossible to achieve a better approximation ratio even if one just asks for homogeneity. We complete the picture by studying several additional social choice properties; for these properties, we prove that algorithms with an approximation ratio that depends only on m do not exist. In this thesis we consider also the important computational problem of bribery in elections, where the winning candidate is computed using a scoring voting rule. In the bribery problem we have an external agent who wants to change the preferences of some voters to make his favorite candidate win the election given a budget. In this thesis we consider scoring voting rules where the voter gives to the first candidate κ points, λ points to his second most preferred candidate and zero points to all other candidates. We prove that for this class of rules bribery is a computationally hard problem. The class of scoring voting rules includes plurality and 2-approval for which an optimal bribing strategy can be computed efficiently as well as 3-approval which is hard to bribe. Concluding we derive that the class of rules we consider is one of the most simple scoring voting rules that are resistant to bribery.

In a representative democracy the people's representatives are expected to do what the people would do if they were present in person. To attain this ideal requires that the legislature in its composition embodies the politically relevant diversity that exists within society, and that the legislature has power to act. These two requirements are prevalent among significant theories of representation, post- Charter court rulings, and the commonly accepted expectations of the people themselves. Typically, the composition of the BC legislature is not representative; and the legislature lacks power to act. The Single Member Plurality electoral system manufactures majorities in the legislature where none exist among the people. Most voters are not represented in the legislature, and the artificial majorities give cabinet undue power. When cabinet has too much power, the concept of responsible government is subverted, MLAs lose their independence, and are beholden to their political party, instead of their constituents. Replacing the Single Member Plurality system with the Single Transferable Vote has the potential to give voters more choice, waste fewer votes, bring greater diversity into the legislature, lessen party discipline, weaken the power of the Premier and cabinet, increase the power of the legislature, restore responsible government, render government more responsive to changing public demands, reconnect government to the people, and give voters power over their representatives. Our electoral system is designed to benefit political parties - not people. Therefore, change will not likely originate with parties and party activists. It must come from the people themselves, aided perhaps by the courts.

The primary aim of this thesis is to advance economics' understanding of the organization of political parties, for the purpose of explaining the policy choices that result from collective decision procedures. Motivating this inquiry is the benign neglect that the political party as an organization has long suffered from in economics, in a manner that mirrors depictions of the firm in early neoclassical analysis. Accordingly, this thesis first considers the question of the relative influence of different contributors to the political parties' electoral activities, i.e. special-interest groups contributing money and individual party activists volunteering their time, on their choice of policy platforms. It is found that the presence of activists induces parties to offer differentiated policy platforms, even in the presence of a special-interest group whose contributions are perfectly substitutable with those of activists. Concurrently, the special interest's influence is to bias the parties' platforms towards its preferred policy. Second, the internal dynamics of parties organized into factions sharing common goals are investigated. It is studied how they affect the party leader's choice of policies while in office and her accountability to voters, through the threat of her removal from the party's helm. While occasionally acting as a distortion on the election mechanism's effectiveness for keeping politicians accountable, the presence of the politician's party is accountability-enhancing especially in the presence of other distortions. This contributes to a second-best theory of politics. This thesis' secondary aim is to contribute to restoring the use of moral and ethical concerns in normative analysis and political economy. This is warranted by the fact that moral and ethical motives matter more in such contexts than in most market transactions, where rational self-interested behaviour largely prevails. This objective is primarily represented in this thesis' study of normative analysis as conditioned on a societal consensus. This study asks how redistributive policies are to be optimally-chosen when the extent of societal co-operation regarding work participation depends on a social norm. Its main finding is that constraining the social planner's choices on the extent of societal cohesion restricts the scope of redistribution compared with an unconstrained social planner. / Thesis (Ph.D, Economics) -- Queen's University, 2011-11-03 12:16:33.632

This thesis seeks to explore the idea of Indigenous participation in Canada’s federal electoral system and whether such involvement can wield positive change to the Canadian/Indigenous relationship. The analysis presented throughout this thesis highlights the development of a tarnished Canadian/Indigenous relationship as well as the debate surrounding the belonging of Indigenous peoples in relation to the Canadian state and their own Indigenous nations. Additionally, this thesis demonstrates that Indigenous peoples voting in Canada’s electoral system do not hinder Indigenous sovereignty, but may heighten its recognition instead, as well as how participation in Canada’s political system may wield influence by Indigenous peoples over Canada’s elected officials. Although potential for Indigenous influence within Canada’s electoral process exists, further research is needed to delve into the subject matter at a deeper level.

This thesis examines whether the introduction of citizenship education in New Zealand would increase the levels of active citizenship of young migrant New Zealanders, using voter turnout as a measure. This research draws attention to an overlooked part of New Zealand political science research by studying young migrant New Zealanders. The theories and topics covered in this research include the notion of citizenship, education, political participation, social capital and multiculturalism. Both surveys and interviews are used to gauge young New Zealanders’ political attitudes and opinions, and their levels of political knowledge, interest, sophistication and socialisation are examined. The findings of this research suggest that citizenship education would not only be beneficial to young migrant New Zealanders but that citizenship education would benefit all young New Zealanders. It is argued that an increase in all young New Zealanders’ levels of political knowledge and interest are likely to have a positive effect on voter turnout levels. It is also argued that the current education curriculum does not explicitly encourage such learning. Using these findings in conjunction with theory, it is recommended that citizenship education is implemented into the New Zealand secondary school curriculum to ensure that New Zealand’s democracy is in good health for the current and future generations.

My thesis develops three models of political economy, examining different factors that affect equilibrium in political markets. The first paper develops a model based on that of Moutos (2001) whereby a government must choose between a tariff and an income tax in order to raise revenue to finance redistribution from rich to poor. I use a simple median voter model of political economy to show that an income tax may be preferred if it can raise more money than the tariff. This result links well with the empirical observation that more liberal trade regimes are often associated with larger government sizes. The second paper explores the idea of interactions between different parts of a political party’s platform and the benefit that different groups can receive from those policies. I show that even when parties have no predisposition towards any particular policy their policy announcements may differ due to the difference in demand for policy favours from special interests. I also discuss how this difference in demand can affect the relative success of interests groups and of the political parties themselves, and apply these results to a simple model of trade policy to show that left-wing parties proposing higher income tax rates may attract support from groups who support trade protection in developed countries. My third paper provides an extension to the well known model of special interest politics by Grossman and Helpman (1996). I introduce costly informative spending that special interests can use to convert uninformed voters into informed ones. This is advantageous to special interests when those being informed are of a similar political persuasion to the interest group members, thus skewing equilibrium policies towards the group’s objectives.

This thesis attempted to determine the validity of the hypothesis that the Jewish vote, as measured by the 1972 election results, was a reflection of the developing importance of economic and social status to the American Jew. By examining the political group behavior that Jews have exhibited in the past and during the 1972 election, the contention of this thesis was to depict a particular, it examined the relationship between the economic level of the Jewish voter, and a switch to the Republican Party in the 1972 election.Steadily Democratic since the 1930's, this "ethnic Jewish vote" had been unaffected by the social and economic advances made by the Jewish people. However, the 1972 election proved that this vote was not immovable. The factors that entered into this relatively large exodus from the Democratic Party were also examined, as were its implications to the present, and predictions for the future of the Jewish vote.

Concerned with explanation and prediction of legislative decision-making, this thesis was based on the premise that relatively structured, discernable patterns of behavior underlie the everyday decisions made by legislators. It has focussed on the final stage of legislative decision-making, the recording of preferences on roll call votes, and has tested the hypothesis that legislators form voting groups as they take voting "cues" from each other.Using "Q" factor analysis, the roll-call voting of the 1971 Session of the Indiana Senate was examined and five voting groups were found. The voting groups were named according to casual inferences made by the researcher.The implications of the findings of this thesis were discussed and specific suggestions for further research were presented.

Previous research shows that women are more likely to vote Democrat than men. Using the 2000 Middletown Area Survey this paper tests the Feminist Consciousness Theory as a possible explanation for the gender gap in voting. Results indicate that women in the study voted more Democrat than men. Those who scored higher on the NonTraditional Gender Role Ideology scale, the Support for Gender Equality Scale and who said that they were a feminist were more likely to vote Democrat. Women were more likely to support gender equality and identify as being a feminist than were men. This means that because women are more likely to have a feminist conscious they are more likely to vote Democrat. / Department of Sociology