In developed countries, the evidence on the impact of school type on student performance is mixed. Researchers are also interested in finding out the effect of private school competition on educational outcomes. The evidence on this for developed countries is mixed as well. What is the effect in developing countries? There are not sufficient studies for developing countries to reach one conclusion. Using data from the survey of the Ministry of Education, Nepal-2005 for School Leaving Certificate Exam (SLC), this dissertation attempts to seek answers to the above two issues for the case of Nepal. The first part of this study analyzes private and public school performance using OLS and logistic models. The study adopts the propensity score matching technique to account for the selection bias problem. The second part of this dissertation attempts to explore the impact of private school competition on public school performance using the number of private schools in the neighborhood as a continuous measure of competition. A binary measure of competition is also used where school is defined to face competition if there is more than one private school in the vicinity of the sample public school. However, in this analysis, there exists an identification problem because private school enrollment is likely to be correlated with public school performance. To address this problem, the study uses the existence of a motorable road within an hour's walking distance from the sample school as an instrument for number of private schools in the neighborhood. The results from the OLS and logistic estimation on the effect of school type on student performance show that public schools consistently have a negative relationship with student performance. On the impact of private school competition on public school performance, the OLS results show no significant relationship using both continuous and binary measures of competition. In contrast, the IV method indicates a positive and significant impact of private school competition on public school performance, which holds true for both types of measures of private school competition.

This dissertation is a comparative study of school discipline in the United States and in Greece. It examines the effect that schools, particularly their organizational form and rules, have upon the behavior of students and how this behavior is understood and categorized. The empirical findings show that, despite facing an elaborate system of rules, punishments, and staff dedicated to discipline, students at a New York school were three times more likely to be unruly compared with students in a similar school in Athens, where only teaching staff managed behavior, and formal rules and regulations governing student conduct were virtually non-existent. Drawing upon the theoretical insights of Emile Durkheim, Mary Douglas, Tom Popkewitz, Michel Foucault, Pierre Bourdieu, and neo-institutionalist scholars, this study proposes explanations for this surprising pattern. I argue that increased structural-functional differentiation within schools and heavy-handed sets of rules and punishments for students erode the moral authority of the teacher and create spaces outside the classroom where students can develop and employ identities and cultural hierarchies that lead to more frequent and extreme forms of unruliness. I also argue that the regulation of student discipline is part of the broader system of state regulation and control. In societies where govermentality is a dominant theme, school discipline becomes preoccupied with questions of measurement, care, and efficiency. What is needed, I suggest, is a return to democracy.

What are the socioeconomic implications of the time allocation decisions made by low-income adolescents? The way adolescents allocate their time between schooling, labor and leisure has important implications for their education attainment, college aspirations, job opportunities and future earnings. This study focuses on adolescents and young adults in urban areas of Brazil that, due to household income constraints, family or peer pressures enter the labor market at an early age, stop studying, and/or start engaging into risky behaviors, such as drug use or sexual activities. The key policy question in this context is then: what incentives could prove an efficient tool to change the time allocation patterns and behaviors that make adolescents drop out of school, fall pregnant (or impregnate) or consume drugs? This dissertation uses data from the Young Agent Project (YAP) a Conditional Cash Transfer (CCT) Program targeting exclusively adolescents in Brazil to examine this issue. This program targets adolescents aged 15 to 17 and its goals are to improve the socioeconomic and educational outcomes of youth in Brazil. The research in the dissertation seeks to determine whether the program has indeed influenced or not the time allocation decisions of low-income youth in Brazil, thus improving their socioeconomic and educational outcomes. The research addresses this issue in three different levels of analysis: 1) whether the YAP has affected schooling outcomes, youth labor decisions and risky behaviors, by gender, ethnicity or region, 2) whether transferring cash directly to the adolescent is more efficient than transferring to the parents, on improving schooling, labor and risky behavior outcomes, and 3) Whether the number of hours per week dedicated to the YAP's after school program is a strong predictor of better outcomes. The data used is the 2006 Projeto Agente Jovem dataset, which is a matched non-experimental, with a treatment group and a constructed control group. This dataset is representative of the recipients of the YAP across regions, states, genders and racial composition, which was administered to 2,210 households with adolescents aged 16 to 20 at least one year after having finished the program. For the analysis, this study used econometric techniques such as Propensity Score Matching (Average Treatment Effect on the Treated, Nearest Neighbor with Replacement) and performed robustness checks with a sensitivity analysis by comparing the treatment effects obtained from linear regression and Inverse Probability of Treatment Weighting.

Although the concept of quality has been an essential part of the higher education sector, the global quest for Quality Assessment (QA hereafter) has raised attention to quality to a new level. Van Vught and Westerheijden (1994) argue that the common charateristics of QA policy include: meta-level organization to conduct external evaluation, self-evaluation, peer review (site visits), published reports, and no link to public funding. Based on the document analysis, at least 48 countries established a meta-organization to implement QA between 1983 and 2010. This finding substantiates that QA has become a "global education policy." This term, coined by Verger et al. (2012), refers to "similar education reforms and a common set of education policy jargon [that] are being applied in many parts of the world" (p. 1). Influenced by policy borrowing and lending theory (Steiner-Khamsi, 2012), this dissertation was interested in understanding why a "global education policy" such as QA resonates in Thailand. Furthermore, it explored how the policy was introduced, implemented, and interpreted by the stakeholders within the higher education sector since its inception. This dissertation deployed a qualitative case study methodology with a triangulation of three methods to collect data: document analysis, 80 elite interviews, and a three-month internship at the Office of National Educational Standards and Quality Assessment (ONESQA).In 1994, QA was a topic of policy discussion within Thailand's higher education sector. However, it was not until 1999, with the promulgation of the National Education Act, that QA became a legal mandate for every institution. ONESQA, a public organization, was established in the year 2000 to be responsible for conducting external quality assessment. Comparing the timing of Thailand's institutionalization of QA to the international landscape, the country is considered to be a late adopter of the policy. Steiner-Khamsi (2006) aptly argued that a study of a late adopter is equivalent to a study on globalization.Why did a "global education policy" such as QA resonate in Thailand? It is argued that historical legacies of the Thai state, policy timing, and the policy belief systems of the policy elites help to explain the policy resonance and the introduction and implementation of QA in Thailand. Historically, the Thai elites have always, actively and purposefully, made reference to policies from elsewhere in order to legitimize national reform. The historical development of the Thai higher education sector is the case in point. While the establishment of universities during the early days of nation-building relied on European models, American models became the blueprint after the Second World War. QA resonated in Thailand because of this historically rooted logic of the Thai state that favored global trends. Policy timing was also an important factor. Given that QA was a byproduct of a sector-wide education reform, the Asian Economic Crisis provided a much needed "window of opportunity" to legitimize the new legal framework (Fry, 2000). The policy belief systems of Thai policy elites mattered. Through cross-sectoral borrowing, senior bureaucrats, university executives, and representatives from the private sector based QA on their professional experiences.How was QA introduced, implemented and interpreted within Thailand's higher education sector? The introduction of QA was perceived as an "irresistible global model/ trend" that Thailand needed to follow. The implementation of QA has been amalgamated within the existing hierarchy, seniority, and structure of Thailand. While QA policy was bureaucratized in every visited university through the creation of a central office, there has also been a professionalization of QA at extensive meetings, trainings, data collection, and self-assessment reports. The interpretation of QA differs depending on the type of institution, official position of each individual, and academic discipline. It is evident that university executives are more favorable to QA, while academics criticize QA for its abundant paperwork and question the link between quality education and quality assessment.This dissertation primarily contributes to the advancement of policy borrowing and lending theory. In addition to the politics and economics of borrowing, the case of Thailand elucidates the existence of a culture of borrowing. Thailand deploys externalization strategy to justify the locally and historically rooted logic and aspiration that becoming modern and adapted to global trends is a national necessity. By identifying the multiple faces of globalization and how it has been used as problem, policy, and politics, this dissertation also contributes to strengthening Comparative Policy Studies.

Traditionally youth policy has been studied from a psycho-social perspective that treats the concept of youth as a natural developmental stage. This dissertation adopts a political perspective and analyzes how political actors shape the social construction of youth. It documents the extent to which five sub-issues of youth policy--education, criminal justice, public health, social services, and workforce development--were present on the congressional agenda from 1973-2008. This research question is addressed through an analysis of congressional hearing data from a researcher-designed database of all congressional hearings held on youth-related issues during this 35 year-period (n = 986). This descriptive analysis provides a longitudinal picture of what Congress chose to consider with regard to youth issues. The dissertation then empirically probes possible explanations as to why the five sub-issues of youth policy were more or less prevalent on the congressional agenda. Drawing from existing literature, this research posits two competing theories that may explain congressional attention to youth issues over time. The external events hypothesis argues that youth issues are present on the agenda as a result of external events catalyzing an increase in attention to youth issues, whereas the internal actors hypothesis asserts that internal actors such as congressional leaders and interest groups are responsible for promoting youth issues. These competing explanations are then tested with a content analysis of the hearings, supplemented with data from a small number of elite interviews. Results suggest that both hypotheses are partially correct, but that the first theory better explains the peaks in the number of hearings, signifying the role external conditions played in motivating Congress to hold more youth hearings.

Since the mid-1990s, public education provision in Turkey has been in constant transformation, a result of modernization efforts connected to the political determination of governments to complete Turkey's accession to the European Union. During this period two nation-wide reforms stand out due to their dramatic impact on children, students, teachers and the education system as a whole. First, the Compulsory Education Law enacted in 1997 required that all the children enrolled in grade 4 or lower must stay in school until the completion of the eighth grade. Second, in 2002, the Ministry of National Education (MONE) abandoned recruiting teachers based on lottery and started to use teachers' test scores instead. Following new legislation, the Center of Measurement, Selection and Placement (ÖSYM) launched a central examination process which is known as the Public Servant Selection Examination (KPSS). This dissertation provides an econometric evaluation of the impact of these interventions on education outcomes in Turkey. The dissertation seeks to establish a causal link between the enactment of KPSS and student achievement. It presents evidence indicating that teacher recruitment via a meritocratic, test-based assessment instead of a lottery may have a positive impact on student achievement. The research also shows that the increase in the average student achievement displayed by Turkey in international assessments such as PISA (Programme of International Student Assessment) and TIMSS (Trends in International Mathematics and Science Study) may be partially explained by the inception of KPSS. The identification strategy for this assessment is based on the fact that the TIMMS data includes information on teachers and test scores for each student sampled in Turkey in both1999 and 2007, that is, before and after KPSS was enacted in 2002. This allows the estimation of a difference-in-differences model with student fixed effects. The findings highlight that students whose teachers were recruited after the enactment of KPSS perform 0.2 standard deviations higher than their counterparts whose teachers were recruited before the enactment of KPSS. This finding remains stable in several sensitivity and robustness checks.The dissertation then turns to analyzing an earlier intervention, the Compulsory Education Law of 1997. The research estimates the impact of the Compulsory Education Law on the years of schooling of women aged between 18 and 29. For this purpose, the dissertation uses the Turkey Demographic Health Survey 2003 and 2008. The identification strategy is based on the fact that, first, cohorts born after 1986 (children enrolled in grade 4 in the1996-1997 school year and later) were subject to the Compulsory Education Law and earlier cohorts were not, and, second, the intensity of the intervention varied between regions. Hence the investigation exploited the between-cohort and between-region variation in intensity of the intervention to estimate the causal impact of the Compulsory Education Law on years of schooling. The findings suggest that the Compulsory Education Law led to a 34 percentage point increase in the probability of completing eight years of schooling and an additional 1.5 years of schooling. Also, the econometric results indicate that the Compulsory Education Law affected high school completion rates, i.e. eleven years of schooling. The analysis of the impact of the Compulsory Education Law is extended to a two-stage least-squares (TSLS) estimation of the impact of completing eight years of schooling/additional years of schooling on teenage marriage and fertility. The between-cohort and between-region variation in intensity of the intervention are used to instrument completing eight years of schooling and additional years of schooling. However, in contrast with the existing research on this issue in Turkey, these TSLS estimations did not supply any evidence in favor of the presence of a causal link between completing eight years of schooling/additional years of schooling and teenage marriage and fertility.

This study explored the skill sets possessed by speech-language pathologists working with profoundly deaf children in three types of settings (state-funded "4201" schools for the deaf, Board of Cooperative Educational Services programs, and local school districts) throughout New York State. The phenomenological method of inquiry was utilized to investigate these skill sets within the varying contexts of speech pathologists' work environments and the deaf students within those settings. Fourteen speech-language pathologists were interviewed for this study; data was triangulated by the collection of responses to case studies and field notes. When possible, supervisors of participants were interviewed as well. Results demonstrated that speech pathologists working with the deaf population possess numerous specialized skills, to varying degrees, depending on the context within which they practice. Findings have implications both for clinical preparation and practice, as well as for education planning and policy in New York State.

Mathematics education in the elementary schools has experienced many changes in recent decades. With the curriculum becoming more complex as a result of each modification, immense pressure has been put on schools to increase student proficiency. The Common Core State Standards is the latest example of this. These revisions to the mathematics curriculum require a comprehensive understanding of mathematics that the typical elementary teacher lacks. Some elementary schools have begun changing the organization of their classrooms from self-contained to departmentalized as a possible solution to this problem. The purpose of this quantitative study was to examine the effects of elementary departmentalization on student mathematics proficiency. This was done by exploring and comparing the background and educational characteristics, teaching practices, assessment methods, beliefs, and influence of departmentalized elementary mathematics teachers. The study also investigated the circumstances under which there are significant differences in mathematics proficiency between departmentalized and non-departmentalized elementary students, and examined if these differences continued into students' eighth-grade years and/or led to higher level eighth-grade mathematics course attainment. Additionally, the study aimed to determine if there was a relationship between elementary departmentalization and mathematics proficiency and also to identify additional factors that could lead to mathematics proficiency. Data came from the U.S. Department of Education's Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K) data set. The ECLS-K is a national data set that followed the same children from kindergarten to eighth grade focusing on their school experiences from 1998 to 2007. Numerous statistical analyses were conducted on this rich data set, utilizing the statistical software Stata 13 and R. The results of this study indicate that there is a significant difference in the mathematics proficiency of departmentalized and non-departmentalized students when teachers have below-average mathematics backgrounds. The students of the mathematically below-average departmentalized teachers displayed the highest mathematics proficiency as well as the biggest gain in mathematics proficiency, and these higher proficiencies and gains continued into later grade levels. However, when exploring differences in mathematics proficiency among all students, there were no conclusive differences between departmentalized and non-departmentalized students. Regression models yielded inconclusive results as well, even after controlling for factors pertaining to classroom size, student demographics and socioeconomic status, student confidence, parental background, teacher knowledge and instructional practices, and prior student mathematical proficiency. Other findings include self-contained and departmentalized third-grade teachers being very similar in their educational backgrounds and teaching practices, whereas departmentalized and non-departmentalized fifth-grade teachers were found to be fairly different in their educational backgrounds and instructional practices. However, in both grade levels, self-contained teachers appeared to be more reliant on printed materials than departmentalized teachers.

Research suggests that suspensions are detrimental to students' socio-emotional and academic development and to the likelihood that they graduate from high school. This literature often describes the types of students who are suspended or the relationships between suspension and student outcomes, either through qualitative methods or quantitative methods that fail to adequately account for the differences between students who are and are not suspended. Using longitudinal administrative data from New York City, I build upon extant research by estimating the link between suspension and short-term academic outcomes within a student fixed effects framework. This approach eliminates unmeasured differences across students that are associated with both the likelihood of being suspended and important student outcomes. I then estimate the link between suspension and long-term academic outcomes using propensity score matching. I find that suspension is associated with increased absences and latenesses, and a decreased likelihood of passing courses the term in which the suspension occurred. Furthermore, suspension is associated with a lower likelihood of graduating within four, five, or six years, and a decreased likelihood of passing Regents exams.

This study aims to explore the impact of college quality on early labor market outcomes in China, including the fresh college graduates' initial employment status and starting wages for students who graduated in 2011. The main data source is the College Student Labor Market (CSLM) survey conducted by Tsinghua University. Distinguished from previous Chinese studies that merely utilized the broad and abstract college quality categories to measure college quality in China, input-based school resource indicators, including faculty-student ratio, proportion of faculty members holding doctoral degrees, average freshman National College Entrance Examination (NCEE) score, and teaching expenditure per student are collected to measure college quality in China for the first time. To identify the causal effect of college quality, the instrumental variable approach and the propensity score matching method are employed to account for the endogeneity of elite college attendance in addition to the traditional ordinary least squares (OLS) regression. To explore the heterogeneous effect of college quality varying by student and family background characteristics, a series of interaction terms are generated in the OLS regressions. The quantile regressions are employed to explore the effect of college quality varying by earning distribution. Moreover, the Heckman correction approach is used to test for potential sample selection bias. This study finds solid evidence that elite college attendance generally has a positive and statistically significant effect on the initial employment status and starting salaries for fresh college graduates who intend to work after college graduation. I find weak support for the existence of heterogeneous effect of college quality. Less-capable students tend to benefit more from attending elite colleges. However, the impact of college quality does not seem to vary by graduates' earning distribution. When using the input-based college quality measures, the results suggest that the quality gap does exist between elite and non-elite colleges in China and the major finding that there is a positive impact of college quality on the starting salary still holds. Some input indicators have stronger correlations with college graduates' starting wages than others.