HOW DOES PUBLIC SPENDING IMPACT EMPLOYMENT OUTCOMES? Evidence from Indian public investment into the technical tertiary education sector

印度技术毕业生的就业结果（如工资/小时工资）一直不达标，但地区教育预算对就业结果的影响历来没有得到充分的实证研究。此外，与西方经济体相比，主流出版物尚未对印度教育预算对经济长期效应的影响和效率进行深入研究。本文旨在利用国家高等技术教育支出的数据，使用不同的统计方法研究队列和国家效应对工资和工作时间的影响，从而弥合知识差距。本规范将包括固定效应回归模型以及在实证研究中很少尝试的连续差分模型。我们还建议对与城乡住区、宗教和种姓有关的不同亚人口进行异质性分析，以探讨预算分配是否公平。该分析还将包括平权行动政策的变量。通过初步回归分析，我们发现，虽然落后的国家预算对当前员工的工作时间有积极影响，但对当前工资和小时工资有负面影响，而且这种负面影响对农村人口和低种姓公民更为明显。这些结果为利用预算的具体方面进行更深入的研究开辟了可能性；例如，与向行政或基础设施提供资金相比，向教学或奖学金提供资金是否会带来更好的就业结果。这篇论文还具有重要的政策含义，并有可能揭示可能的预算效率低下和无意的党派分配效应。论文的结果可以传达关于弱势群体是否受益的重要信息，或者是否存在挤出效应（更有特权的亚群体挤占了技术教育的入学和工作机会）和/或奶油层问题（即，平权行动政策导致少数群体中的富裕学生取代非少数群体中的贫困学生的担忧），平权行动政策被忽视了。

Employment outcomes, such as salaries/hourly wages, have beensubstandard among India's technical graduates, but the impact of regionaleducation budgets on employment outcomes has historically been empiricallyunderstudied. Further, in comparison to Western economies, a thorough studyof the impact and efficiency of the Indian education budget on economiclong-run effects has not been done in mainstream publications. This thesisaims to use data on state tertiary technical education spending to study theimpact on salaries and work hours with cohort and state effects using differentstatistical methodologies, thereby bridging the knowledge gap. Thisspecification will include both fixed effects regression models as well as acontinuous difference-in-difference model that has seldom been attempted inempirical studies. We also propose to conduct a heterogeneity analysis withdifferent sub-populations relating to rural/urban settlements, religion, and casteto explore whether the budget allocations are equitable. This analysis will alsoinclude a variable for affirmative action policies. Through preliminaryregression analysis, we find that, while the lagged state budgets had positiveeffects on current employee work hours, they had negative effects on currentsalaries and hourly wages, and this negative effect was more pronounced forrural populations and lower caste citizens. These results open up thepossibilities for more in depth research using specific aspects of the budget; forinstance whether funding into teaching or scholarships lead to betteremployment outcomes as opposed to funding into administration orinfrastructure. This thesis also has important policy implications and has thepotential to shed light on the possible budget inefficiencies and unintendedpartisan distribution effects. The results of the thesis can convey importantinformation on whether disadvantaged populations are benefited or if there is acrowding out effect (more privileged sub-populations crowding out admissionsand job opportunities in technical education) and/or creamy layer problem (i.e.,the concern that affirmative action policies result in wealthy students withinminority groups displacing poorer students in non-minority groups) thataffirmative action policies are ignoring.