iStock

Interested in contributing to our mission of advancing immigration policy worldwide? Read on for more details about our new position.  


About IPL

The Immigration Policy Lab (IPL) at Stanford University designs and evaluates policies to advance the integration of immigrants and refugees worldwide. By producing new evidence and translating it into creative solutions, we seek to improve refugees’ and immigrants’ opportunities and strengthen their host communities.

Using new data and cutting-edge analytical tools, we bring evidence to bear on the urgent problems facing immigrants, refugees, and their service providers. We engage with community-based organizations as well as local, state, national, and international government agencies to test the effectiveness of current policies. In addition, we co-design new policies, programs, and tools that ultimately affect millions of immigrants, as well as increase the economic and social prosperity of the communities in which they live. Our talented team of faculty, professional staff, postdocs, and students has created a research model that combines the quality and rigor of an academic lab with the efficiency and innovation of a civic-tech startup.

About the Position

Stanford University is seeking to appoint four Postdoctoral Research Fellows as part of a cluster hire to work on its GeoMatch research portfolio at the Immigration Policy Lab. In partnership with governments, these projects study and implement an algorithm-based matching tool to connect refugees and immigrants to locations within a host country where they are most likely to thrive. GeoMatch is a global research initiative that combines high-quality academic research with the work of a tech startup: product design, strategic partnerships, and advances in artificial intelligence.

About the Project

The place where immigrants settle within a host country has a powerful impact on their lives. This destination can be a stepping stone and provide opportunities to find employment, maximize earnings, learn the host country language, and access services such as education and healthcare. Location decisions therefore not only affect immigrants themselves, they also shape immigrants’ contributions to the local economy and society. This project seeks to develop and test data-driven matching tools (called GeoMatch) for location decision-makers—both governments and immigrants themselves—that generate personalized location recommendations, leveraging insights from historical data and human-centered AI. The goal is to advance both the theoretical and empirical frontiers of algorithmic matching for newcomers. On the theoretical front, our interdisciplinary team of faculty experts will tackle problems at the intersection of estimation and prediction, algorithms and mechanism design, human-AI interaction, and immigrant integration. On the empirical front, we plan to conduct pilot tests via randomized controlled trials on the use of GeoMatch in collaboration with partners.

About the Opportunity

We are looking for four postdoctoral fellows to join our GeoMatch team and help grow our exciting portfolio of projects related to statistics, machine learning, online algorithms, and immigrant integration around the globe under the supervision of an interdisciplinary faculty leadership who have recently been awarded the Hoffman-Yee Grant at the Human-Centered AI Institute (HAI) at Stanford University. Each postdoc will work on a specific topic area under the supervision of the corresponding faculty leads and focus their postdoc research in that topic area under faculty guidance. The postdoc will have the opportunity to co-author papers envisioned to appear in top journals that report on the results of the studies, work with an array of affiliated faculty from top institutions and develop independent projects related to machine learning in the immigration policy evaluation context. The postdoc positions are:

1. Statistical Methodology and Estimation

2. Online Allocations and Matching

3. Social Choice and Market Design

4. Refugee Integration

The initial appointment will be for one year beginning as soon as possible in fall 2022, with possibility of renewal. Salary is based on experience. Benefits are provided.

General Qualifications: The Postdoctoral Research Fellows will have completed a Ph.D. For specific qualifications, please see each post’s details below. Stanford University is an equal opportunity employer. It welcomes nominations of, and applications from, underrepresented groups, as well as others who would bring additional dimensions to the university’s research missions.

Application Instructions

Applications should be submitted here and include a CV, a cover letter specifying the topic area to which you are applying, along with an explanation of your interest and qualifications for the role; graduate school transcripts; a writing sample; and at least two letters of recommendation. Additional papers may be requested at a later date. Applicants should be prepared to complete a research design and data analysis exercise as part of the interview process.

Applications will be reviewed on a rolling basis, so we encourage applications as soon as possible. Applicants who have questions about IPL may contact the Faculty co-Director of the Immigration Policy Lab, Jens Hainmueller, directly (jhain@stanford.edu Subject: GeoMatch Postdoc Opportunity). Inquiries about specific research should be directed to the faculty listed in the specific position description, but please do not send your application materials to them.

Specific Position Details

Position 1: Statistical Methodology and Estimation

Faculty advisors: Dominik Rothenhäusler (Stanford University), Kirk Bansak (UC Berkeley), and Jens Hainmueller (Stanford University)

Position overview: GeoMatch relies on the ability to estimate refugees’ expected outcomes in different locations prior to their arrival/assignment. Therefore, theory, techniques, challenges, and other topics in estimation and statistical inference are a core part of the GeoMatch research and implementation agenda. Prior and Ongoing Work: The research team has made progress in core research that has helped build the foundation for GeoMatch. This includes the estimation of expected outcomes across locations as a function of refugees’ pre-arrival background characteristics. In addition, we have demonstrated the feasibility of data-driven, outcome-based matching with administrative data sets across multiple countries; investigated the efficacy of different methods for estimating expected outcomes; established the formal conditions under which these estimates provide valid counterfactuals; and performed analyses demonstrating such conditions within real-world asylum and refugee resettlement contexts.

The successful candidate will be expected to work on novel methodology to address various challenges in the modeling and prediction of refugees’ outcomes. This will include methodology on distribution shift and transfer learning, robust machine learning, and/or uncertainty quantification for complex models.

Preferred Qualifications: Ph.D. in statistics, machine learning, or a related discipline. Knowledge or experience in the areas of causal inference, transfer learning, data fusion, or distributional robustness are preferred, but not essential.

Position 2: Online Allocations and Matching

Faculty advisors: Yonatan Gur (Stanford University) and Elisabeth Paulson (Harvard Business School)

Position overview: The implementation of algorithmic approaches for refugee resettlement subject to a variety of constraints, including capacity, locality fairness, and group fairness, presents new foundational and practical operational challenges. The proposed research topics in this cluster span AI-human interaction, fairness and equity in AI, and new practical considerations not addressed by existing research.

A successful candidate will be expected to work on the design and implementation of online resource allocation and stochastic matching algorithms designed to prescribe the geographic allocation of refugees/asylum seekers under capacity constraints. Additional complexities include incomplete information, fairness considerations, and more. Work may include theoretical components such as developing new data-driven algorithms and evaluating their performance guarantees, evaluating practical approaches through analysis of real data sets, and implementation of new successful algorithms.

Preferred Qualifications: PhD in Operations Research, Computer Science, or related disciplines. Expertise in the areas of online decision-making algorithms for matching problems, resource allocation problems, multi-arm bandits, and sequential stochastic optimization.

Position 3: Social Choice and Market Design

Faculty advisors: Avidit Acharya (Stanford University)

Position overview: There are limitations to matching done purely for the sake of improving outcomes. This approach does not include refugees’ preferences, nor does it incorporate private information refugees may have about what host community would work best for them. Relevant literature in market design has focused on purely preference-based matching schemes, where agents are matched to locations or items based on their preferences in situations such as school choice or medical residency. Despite the normative appeal of providing agency and incorporating private information, the purely preference-based approach assumes that agents are well informed in their preferences. This may not be true of refugees, who tend not to be familiar with many of the potential locations. Moreover, purely preference-based matching might not yield the best possible assignment, since it ignores the historical data that can help identify the best assignment for a given individual.

The successful candidate will work with the data team at IPL to develop new market design methods and models to address the refugee matching problem. Familiarity with existing market design literature on refugee matching is preferred, but not essential.

Preferred Qualifications: PhD in computer science, economics, or political science, with expertise in social choice theory, market design, and/or related fields.

Position 4: Refugee Integration

Faculty advisors: Jens Hainmueller (Stanford University) and Tomás R. Jimenez (Stanford University)

Position Overview: One of the important HAI focus areas is to better understand the human impact of AI technologies. To accurately assess the performance of our GeoMatch algorithms, we first need to better understand how the characteristics of the resettlement location and the refugee interact to create integration successes or failures.

The successful candidate will work with the data team at IPL to conduct studies that examine various important questions related to how the resettlement process affects refugee integration outcomes and host communities. For this the candidate will leverage administrative datasets from various countries that IPL has access to. The candidate will also work on the designs and implementations of randomized control trials of the GeoMatch pilots.

Preferred Qualifications: PhD in political science, economics, sociology, and/or related fields. Familiarity with the quantitative and or qualitative study of immigration and refugee related issues is preferred.