SIEPR offers undergraduates research opportunities throughout the year. Undergraduate Reasearch Assistants (UGRA) work directly with SIEPR faculty on research and data acquisition. UGRAs are funded for part-time research for a maximum of 15 hours per week during the academic quarters or for a 10-week Summer Program.
Both programs are available only to currently enrolled Stanford undergraduates. For the Summer Program, students must be enrolled in undergraduate studies at Stanford both the preceding Spring quarter and the following Fall quarter. If you are interested, please apply early. Applications are reviewed as they are received. For questions about the program, please contact firstname.lastname@example.org.
Summer UGRA Program: June 22 - August 28, 2020
Students should be able to work remotely, if needed, for the entirety of the of the ten week program.
If you have already applied to the Department of Economicssummer RA program, please contact Professor Tendall and let him know that you are interested in these SIEPR positions as well.
SIEPR is sponsoring a ten-week Summer RA program from June 22 - August 28, 2020. Summer RAs receive a stipend for 10 forty-hour workweeks that will be devoted to research work under the discretion of their faculty mentor. Please apply to individual positions using the contact information listed to the right of each open position.
Machine Learning (ML) and Causal Inference Resources for Applied Researchers
Post Date: April 16, 2020
Number of Positions Available: 1
Faculty Mentor: Susan Athey
Project Description: Students will work with Professor Susan Athey and members of the Lab to formulate a curriculum for researchers working with common data science, applied economics, and project management tools. Our aim is to develop a toolkit for applied researchers that influence a broad audience of researchers. The initial audience for the curriculum will be new members of the Lab. However, the resources and tools that we develop will be shared to others at Stanford and beyond.
Conflict and Polarization
Post Date: April 16, 2020
Number of Positions Available: 1
Faculty Mentor: Saumitra Jha
Project Description: This project draws on historical natural experiments and contemporary field experiments to examine how economic approaches -- such as financial innovations or inter-ethnic business relationships-- can mitigate violent conflict and political polarization.
Qualifications: interest in archival research, finance, development economics and comparative politics. Python and STATA are desirable. French language skills would be an asset.Details: gathering and analyzing data from archival sources and surveys, scrape websites and conducting literature reviews.
Fiscal Effects on Economic Growth (FILLED)
Faculty Mentor: Michael Boskin
Project Description: The project encompasses a variety of interrelated issues in the analysis of the effects of taxes, spending and public debt on economic growth.
Qualifications: Preferably Econ major with at least the Introductory Econometrics course. Good work habits, stressing attention to detail, checking and confidence to raise any issues. Some knowledge of Stata desirable.Details: Collecting, cleaning and preparing data for analysis and input into sophisticated econometric modelling. Carefully reviewing literature and preparing reports.
Endowment Risks and Returns (FILLED)
Faculty Mentor: Jeremy Bulow
Project Description: In the 1990s and early 2000s major university endowments employing the "Yale Model" of investing earned superb returns. But over the last decade returns have been significantly lower than a simple index of all US stocks. Defenders of the Yale model now have to depend on two more nuanced arguments --- that on a risk-adjusted basis their portfolios are less risky than a stock index fund, and that while US stocks have done particularly well over the last decade we should be concerned with prospective returns, not past returns. The second argument is surely right; the first requires empirical examination. To pick our own university as an example, Stanford's (undisclosed) expected annual costs of endowment management are similar to total revenue from undergraduate and graduate tuition, room, and board. Acknowledging that many private investors the University hires are highly skilled, are we getting our money's worth? And in a world with negative real interest rates on safe assets what are the prospects for endowment performance that is strong enough to finance annual payouts of 5 percent or more and management expenses of 2-3 percent?
Qualifications: Undergraduate economics major. Some basic computer and statistical skills. Willingness to work hard.Details: The initial phase will be to gather information about the portfolio composition of universities, both in aggregate and for the major endowments individually. We may consider other non-profit endowments as well. Ideally we will also be able to find some documents discussing both the estimates of variance and covariance used by endowments in estimating portfolio risk, and hopefully some information about how these estimates are developed. Next will be a literature review on returns in industries like private equity, hedge funds, commodities, and real estate. The principal task will be to translate the returns in these sectors, which are not marked to market, to valuations that would be comparable to those used in the public markets. This may be done in part by looking at whether the reported returns in these sectors can be forecast using past market returns. More difficult, we will attempt to estimate the correlations across sectors to estimate the portfolio risk of these endowments. The project is just starting so work this summer will be early stages.
Using Remote Sensing Data and Machine Learning to Measure Economic Development (FILLED)
Faculty Mentor: Marshall Burke
Project Description: The increasing availability of remote sensing data --- particularly in the form of high-resolution, high-frequency satellite imagery --- presents unique opportunities to study the causes and consequences of economic development. By combining these new data sources with a convolutional neural network (CNN), we have been able to produce highly granular measures of household wealth and consumption for all of sub-Saharan Africa. The proposed project will extend this work in two ways.
First, understanding whether model accuracy is correlated with potential confounders is a critical issue in evaluating how to incorporate the CNN estimates as predictors in causal models. While we have measured the overall prediction accuracy of the CNN estimates, we do not yet know whether the prediction error is correlated with other features that may affect, and be affected by, economic development. Such features include temperature, elevation, the presence of armed conflict, and refugee flows among others. A key goal of the project will be to understand the strength of these correlations and to develop methods to overcome the inferential issues they pose.
Second, we will be utilizing the CNN estimates to revisit three key results in the political economy of development literature. Specifically, we will be re-evaluating core findings on the relationship between an ethnic group's representation in government and changes in levels of development; between improvements in local public goods and electoral outcomes; and between the colonial-era partitioning of ethnic groups and modern-day development. The fine-grained nature of the CNN estimates allow us a unique opportunity to overcome data limitations in each of these important literatures. By the end of the summer, we aim to have preliminary findings for each of the three applications mentioned above as well as an evaluation of their robustness.
Qualifications: Research assistants will need a working grasp of Python and packages relevant for geo-spatial analysis, particularly GeoPandas. They should also have some experience with cloud computing and have some facility in training neural networks and evaluating their performance. Some training in causal inference and advanced econometrics is desirable but not required. Students should have a broad interest in economic development and in the use of remote sensing data to answer important questions in political economy.Details: The research assistants will work closely with Marshall Burke and his lab members in both dataset production, spatial analysis, and the training of conventional and convolutional neural networks with other remote sensing data. Primary responsibilities will be in data pipelining, including the collection of new ground truth data and gathering the corresponding remote sensing data, as well as training CNNs on different configurations of ground truth data. Research assistants will also evaluate changes in the prediction accuracy across time and space, and determine what other features are correlated with model residuals. They will also work on comparing government estimates of wealth to equivalently aggregated estimates produced by the CNN. Finally, they will be required to participate in twice-weekly research group meetings convened by a post-doc in Marshall's lab. When necessary, students will also be asked to attend teaching sessions designed to teach skills necessary for various aspects of the project.
The Impact of the ACA on Hospital Finances (FILLED)
Faculty Mentor: Mark Duggan
Project Description: The Affordable Care Act (ACA) substantially increased health insurance coverage in the U.S. The financial effects for hospitals were potentially quite significant, as many of their previously uninsured consumers now have coverage through the Medicaid program or through private ACA exchanges. In this project, we will estimate the effects of this legislation on hospital balance sheets and how this varied with respect to each state's Medicaid expansion status and the pre-ACA characteristics of hospitals' patients.
Qualifications: Some familiarity with programming and interest in data analysis.Details: Obtaining, processing, and cleaning hospital financial data. Linking this to data on the characteristics of each hospital market and of each hospital's patients. Analyzing how hospital finances changed as a result of the ACA.
Work on book titled "Innovation and Contract: Adaptability and Contract in an Uncertain World” (FILLED)
Faculty Mentor: Ronald Gilson
Project Description: Rapidly innovating industries are not behaving the way theory expects. Forty years ago work by Oliver Williamson that led to a Nobel Prize and became conventional industrial organization theory predicted that when parties in a supply chain have to make transaction specific investments – investments that will have a significantly lower value if they have to be subsequently redeployed outside the transaction – the heightened risk that their contracting counterparty will take advantage of them would drive them away from contractual relationships and toward vertical integration The pressure toward vertical integration would be especially powerful in rapidly innovating industries where swift technological change renders contemporary contract theory unable to resolve the tension between the need for investment in the parties relationship in the face of uncertainty.
Despite conventional industrial organization theory, however, contemporary practice has moved away from vertical integration. Producers today recognize that they cannot themselves maintain cutting-edge technology in every field required for a successful product. Rather than vertical integration, we observe vertical disintegration. In the process, firms by necessity are developing forms of contracting beyond the reach of existing contract theory models. How do firms contract when the uncertainty that results from the speed of innovation makes it impossible to set out what product is feasible to produce, what product specifications will be, or what the price for the effort should be?
This book explores how the escalating uncertainty confronting business today influences what we call the design space for contract: how do contracting parties adapt and collaborate in a rapidly changing business environment? Put simply, we show that in this environment the design space is mapped by the intersection of the level of uncertainty parties must confront and the scale – the number of parties engaged in the relevant activity – in which they will engage.
Qualifications: Mid-level economicsDetails: Research and memo drafting on areas covered in the manuscript, for example, research on the auto industry and efforts to assure the safety of the leafy vegetable food chain.
COVID-19 and Elections (FILLED)
Faculty Mentor: Andrew Hall
Project Description: We will be collecting data to understand the effect of COVID-19 on election administration and election outcomes, to propose policies to ensure that the 2020 elections can be administered properly, and to understand more generally government failures and successes around COVID and how those are related to issues of electoral accountability.
Qualifications: Positive attitudeDetails: Locating data sources online and in archival form, digitizing data, entering data, writing code for web scraping, participating in weekly research meetings, performing literature reviews, writing memos, etc.
Pages of Prejudice (FILLED)
Faculty Mentor: James Hamilton
Project Description: I'm currently working on a book entitled "Pages of Prejudice: How Classifieds Made Media Profitable, Work Accessible, and Opportunities Unequal." It focuses on three questions: How have media served the demand for prejudice in hiring? What are the impacts of this form of media bias? Why does this matter today? The book is a history of the media told through a single market transaction, the sale of information that connects a worker with a hirer. It starts with America's first newspaper, progresses through the golden age of journalism, describes the evolution of job sites and social media platforms, and ends with a discussion of how AI may improve matching of workers with opportunities.
Qualifications: The work will involve searching databases, coding samples of ads, doing summary statistics. This could be done in spreadsheets. An interest in media, discrimination, tech, economic history, and/or public policy would make this a good opportunity to learn research skills. A student early in their economics career might be a better fit, since there will not be highly technical statistical work.Details: The RA will work extensively with online archives of newspapers, including ProQuest Historical Newspapers and America's Historical Newspapers. Topics to explore will include how workers and positions are described by race, gender, ethnicity, and age; how classifications of work changed over time; how opportunities within specific cities were offered or foreclosed by limitations described in classifieds. Additional work may entail studying how current job sites classify opportunities and workers. There may also be research involved about the evolution of public policies affecting job advertising, ranging from civil rights legislation to recent law suits about Facebook's use of gender and age in employment ad targeting.
The Future of Clean Water Act Enforcement (FILLED)
Faculty Mentor: Daniel Ho
Project Description: The Regulation, Evaluation, and Governance Lab is working on an ambitious research project to envision the future of environmental enforcement based on machine learning and to understand the current constraints on bureaucratic capacity. The project includes developing computer vision models for Clean Water Act violations; evaluating the impact of intensive livestock farming on environmental and health outcomes; and implementing a field trial with EPA and partner states to use risk models in enforcement.
Qualifications: We are looking for intellectually curious students with an interest in using their skills to solve some of the most complex challenges facing society today. Interested students should submit their resume, transcript (unofficial is fine), and cover letter to email@example.com. Some exposure to machine learning and/or computer science is a plus.Details: We seek research assistants to help collect data, conduct literature reviews, develop a web app, and implement statistical analyses.
Regulation of Financial Advisors (FILLED)
Faculty Mentor: Colleen Honigsberg
Project Description: There are multiple regulatory classifications for what most consumers colloquially refer to as a financial advisor (e.g., a financial advisor may be classified as a broker-dealer, investment adviser, etc.). The classification is determined based on the type of advice the advisor provides to his clients, and the type of products in which he typically transacts. However, the different classifications have significant overlap, making it easy for advisors to move from one classification to another. The ease with which advisors can move from one regulatory regime to another—and the significant differences across regulators—creates incentives for “bad” advisors to target the least stringent regulator. My paper examines this potential phenomenon.
Qualifications: The ideal RA would be familiar with Excel, Stata, and Python. More importantly, the RA must be detail-oriented as the RA will need to become familiar with the regulatory environment and relevant institutional detail to manage the data properly.
Details:The RA will be responsible for collecting and managing data from several different regulators. Most of the data will be received directly from the regulators through Freedom of Information Act requests, but some data will need to be scraped from different government websites. The different regulators are likely to provide data in different formats, so all files will need to be cleaned and standardized. Further, the data may need to be merged using a fuzzy match.
Stanford Intellectual & Developmental Disabilities Law & Policy Project (FILLED)
Faculty Mentor: Alison Morantz
Project Description: The Stanford Intellectual & Developmental Disabilities Law & Policy Project (SIDDLAPP) publishes academic research and policy reports around legal and economic policy issues that affect the I/DD community in California and nationwide. For more information on the project's goals, resources, and publications, see https://law.stanford.edu/siddlapp/.
Qualifications: Although there are no prerequisites for the position, all of the following qualifications are highly desirable: