Search

Education is a fundamental input of human capital formation. In this dissertation we explore topics related to how much and what time of human capital individuals invest in, and the long term-consequences of these investments. We begin with by measuring the degree to which financial incentives can affect a college student’s field of study. Next, we attempt to identify genetic variants associated with increased educational attainment and examine the biological systems implicated by this analysis. Last, we test for heterogeneous treatment effects of education on health across the distribution of observed health and across a genetic predictor of health. In chapter 1, we examine whether students respond to immediate financial incentives when choosing their college major. From 2006-07 to 2010-11, low-income students in technical or foreign language majors could receive up to $8,000 in SMART Grants. Since income-eligibility was determined using a strict threshold, we determine the causal impact of this grant on student major with a regression discontinuity design. Using administrative data from public universities in Texas, we determine that income-eligible students were 3.2 percentage points more likely than their ineligible peers to major in targeted fields. We measure a larger impact of 10.2 percentage points at Brigham Young University. In chapter 2 we find that, educational attainment (EA) is strongly influenced by social and other environmental factors, but genetic factors are also estimated to account for at least 20% of the variation across individuals. We report the results of a genome-wide association study (GWAS) for EA that extends our earlier discovery sample of 101,069 individuals to 293,723 individuals, and a replication in an independent sample of 111,349 individuals from the UK Biobank. We now identify 74 genome-wide significant loci associated with number of years of schooling completed. Single-nucleotide polymorphisms (SNPs) associated with educational attainment are disproportionately found in genomic regions regulating gene expression in the fetal brain. Candidate genes are preferentially expressed in neural tissue, especially during the prenatal period, and enriched for biological pathways involved in neural development. Our findings demonstrate that, even for a behavioral phenotype that is mostly environmentally determined, a well-powered GWAS identifies replicable associated genetic variants that suggest biologically relevant pathways. Because EA is measured in large numbers of individuals, it will continue to be useful as a proxy phenotype in efforts to characterize the genetic influences of related phenotypes, including cognition and neuropsychiatric disease. In 1972, the mandatory minimum age at which a student could drop out of school in England and Wales was raised from 15 to 16, constraining roughly 15 percent of the student population. In chapter 3, we exploit this discontinuous increase in educational attainment to estimate the impact of education on body mass index (BMI) and diabetes approximately 40 years later. While previous literature found no significant effect of education on health, they were not able to investigate whether these effects vary along the distribution of health outcomes. We are able to detect large effects on BMI in the upper quantiles of observed BMI, as large as 2 BMI points at the 90th percentile of BMI, from a baseline of 35.6. Using a genetic predictor of BMI, we also find that those with higher genetic risk of obesity see smaller reductions in BMI as a result of the increase in compulsory schooling while large reductions are seen in those with low genetic risk. Taken together our results point to the importance of considering heterogeneity when estimating the impacts of education on health., Economics

Prices in financial markets are primarily driven by the interaction of risk and time. The returns to financial assets over long time horizons are primarily driven by fundamental news regarding their promised cash flows. In contrast, short-run price variation is associated with a large degree of predictable, transient investor trading behavior unrelated to fundamental prospects. The quantity of long-run risk directly affects economic well-being, and its magnitude has varied significantly over the past century. The theoretical model presented here shows some success in quantifying the impact of news about future risks on asset prices. In particular, some investing strategies that appear to offer anomalously large returns are associated with high exposures to future long-run risks. The historical returns to these portfolios are partly a result of investors’ distaste for assets whose worth declines when uncertainty increases. The financial sector is tasked with pricing these risks in a way that properly allocates investment resources. Over the past thirty years, this sector has grown much more rapidly than the economy as a whole. As a result, asset prices appear to be more informative. However, the new information relates to short-term uncertainty, not long-run risk. This type of high-frequency information is unlikely to affect real investment in a way that would benefit broader economic growth.

Lymph nodes (LNs) are dynamic organs responsible for providing a supportive and centralized environment for the generation of immune response. Utilizing a highly organized network of non-hematopoietic stromal cells, the LN serves as the context in which the immune system collects and presents antigen, promotes innate and adaptive immune interaction, and generates protective cell-mediated and humoral immunity. In this way, proper organization and function of the LN environment is a critical component of effective immunity, and understanding its complexity has direct impact on the ability to generate and modulate primary immune response to specific antigens. To this end, the LN architecture, underlying stromal networks, and environmental and cellular responses to influenza vaccination were investigated. Using novel approaches to conduit imaging, details of the collagen network that comprises the LN scaffolding have been integrated into current understandings of LN architecture. The cellular compartment responsible for the maintenance of that scaffolding, fibroblastic reticular cells (FRCs), have been studied using an induced diptheria toxin receptor model. By specifically ablating the FRC population in mice, their role in the maintenance of T cell homeostasis has been confirmed in vivo. More surprisingly, a disruption of the FRC network resulted in a loss of B cell follicle structure within LNs, and a reduction in humoral immunity to influenza vaccination. These findings led to the identification of a new subset of FRCs which reside in B cell follicles, and serve as a critical source of the B cell survival factor BAFF. Turning towards the hematopoietic response to influenza vaccination, a highly unexpected lymph node resident dendritic cell (LNDC) response has been identified following vaccine antigen deposition within specialized sites in the LN medulla. Rapid migration of LNDCs into these sites optimizes exposure of the population to viral antigen, and de novo synthesis of a CXCL10 chemokine gradient by activated LNDCs ensures efficient antigen specific \(CD4^+\) T cell response, and protective humoral immunity - independent of migratory dendritic cell status. Altogether, these studies highlight a highly dynamic, responsive LN environment with direct influence on primary immune response - the understanding of which has broad implications in vaccine biology.

In addition to leukocytes, secondary lymphoid organs are populated by non-hematopoietic stromal cells. This diverse group of cells supports lymphocyte migration and homing, facilitates antigen delivery, and promotes T cell survival. However, there is relatively little known about the specific molecules governing the roles that these cells play in regulating dendritic cell (DC) motility and lymph node architecture. Here, we examine the interaction between two molecules, CLEC-2 and podoplanin (PDPN), that are critical for DC migration and maintaining structural integrity of lymph nodes. Together, these studies identify novel functions of lymph node stromal cells and a unique function for PDPN in the immune system. In response detecting an potentially harmful antigen, DCs in peripheral tissues mature and travel to downstream lymph nodes by following chemokine gradients secreted by lymphatic endothelial cells (LECs) and fibroblastic reticular cells (FRCs) present in the lymph node paracortex. We discovered that, in addition to chemokines, DC migration requires CLEC-2 on DCs, as engagement of CLEC-2 with PDPN, which is expressed by LECs and FRCs, incites DC motility and is required for DC entry into the lymphatics, efficient arrival in the lymph node, and migration along the FRC network within the lymph node. Next, we examined the effect of this interaction with respect to the stromal cell. Through a combination approaches, we discovered that PDPN is a master regulator of contractility in FRCs. The fact that FRCs are contractile cells was previously reported, but our study is the first to identify a function for this contractility: upon blockade of PDPN-mediated contractility, lymph nodes became enlarged, the FRC network became more sparse, and there were increased numbers of lymphocytes in the lymph node. Importantly, during an immune response, these changes resulted in more proliferation of antigen-specific T cells and impaired contraction of the lymph node upon resolution of inflammation. Finally, we found that CLEC-2 binding PDPN recapitulated the effect of PDPN deletion. Thus, during an immune response, CLEC-2+ DCs would use PDPN to efficiently migrate to the lymph node and simultaneously cause FRCs to relax and prepare the lymph node for expansion.

Non-hematopoietic stromal cells (SCs) promote and regulate adaptive immunity through numerous direct and indirect mechanisms. SCs construct and support the secondary lymphoid organs (SLOs) in which lymphocytes crawl on stromal networks and inspect antigen-presenting cells for surface-display of cognate antigens. SCs also secrete survival factors and chemotactic cues that recruit, organize, and facilitate interactions among these leukocytes. They influence antigen access by secreting and ensheathing extracellular matrix-based conduit networks that rapidly convey small, soluble lymph-borne molecules to the SLO core. Furthermore, lymph node stromal cells (LNSCs) directly induce \(CD8^+\) T cell tolerance to peripheral tissue restricted antigens and constrain the proliferation of newly activated T cells in these sites. Thus, stromal-hematopoietic interactions are crucial for the normal functioning of the immune system. LNSCs are extremely rare and difficult to isolate, hampering the thorough study of their biology. In order to better understand these stromal subsets, we sorted fibroblastic reticular cells (FRCs), lymphatic endothelial cells, blood endothelial cells, and podoplanin \(^−CD31^−\) cells (double negative stromal cells; DNCs) to high purity from resting and inflamed murine lymph nodes. We meticulously analyzed the transcriptional profiles of these freshly isolated LNSCs as part of the Immunological Genome Project Consortium. Analysis of the transcriptional profiles of these LNSC subsets indicated that SCs express key immune mediators and growth factors, and provided important insights into the lymph node conduit network, FRC-specialization, and the DNC identity. Examination of hematopoietic and stromal transcription of ligands and cognate receptors suggested complex crosstalk among these populations. Interestingly, FRCs dominated cytokine and chemokine transcription among LNSCs, and were also enriched for higher expression of these genes when compared with skin and thymic fibroblasts, consistent with FRC-specialization. LNSCs that were isolated from inflamed lymph nodes robustly upregulated expression of genes encoding cytokines, chemokines, antigen-processing and presentation machinery, and acute-phase response molecules. Little-explored DNCs showed many transcriptional similarities to FRCs, but importantly did not transcribe interleukin-7. We identified DNCs as consisting largely of myofibroblastic pericytes that express integrin \(\alpha 7\). Together these data comprehensively describe the transcriptional characteristics of four major LNSC subsets isolated from resting and inflamed SLOs, offering many avenues for future study.