Publications

Forthcoming
Article

Outcome-Driven Dynamic Refugee Assignment with Allocation Balancing

By: Kirk Bansak and Elisabeth Paulson

This study proposes two new dynamic assignment algorithms to match refugees and asylum seekers to geographic localities within a host country. The first, currently implemented in a multi-year pilot in Switzerland, seeks to maximize the average predicted employment... View Details

Keywords: AI and Machine Learning; Refugees; Geographic Location; Employment

Find at Harvard

2016
Working Paper

Through the Grapevine: Network Effects on the Design of Executive Compensation Contracts

By: Susanna Gallani

Effective design of executive compensation contracts involves choosing and weighting performance measures, as well as defining the mix between fixed and incentive-based pay components, with a view to fostering talent retention and goal congruence. The variability in... View Details

Keywords: Compensation Design; Board Interlocks; Compensation Consultants; Network Centrality; Homophily; Quadratic Assignment Procedure; Blockholders; Executive Compensation

SSRN

Research Summary

Statistical Methodology

William Simpson is developing methods of inference to use when assumptions of standard models are not met. He has created a hypothesis test to use for ipsative variables that adjusts for the non-zero correlations among variables expected under the null hypothesis. ... View Details

2007
Other Unpublished Work

Mind Over Matter? Similarities and Differences Between Perceived and Observed Networks

In spite of the rapid development of new methods for network analysis—relying on electronic data sources and sophisticated computational analysis—organizational scholars continue to rely largely on more traditional survey-based methods. We believe that the... View Details

Keywords: Surveys; Organizations; Social and Collaborative Networks

September 2010 (Revised October 2011)
Case

Chances Are? Course Selection at HBS and at Kellogg

By: Hanna Halaburda and Aldo Sesia

The case describes two alternative elective course assignment procedures: Harvard Business School's lottery-based system and Kellogg Graduate School of Management's bidding-based system. The case has been designed to discuss the benefits and drawbacks of each system... View Details

Keywords: Business Education; Curriculum and Courses; Higher Education; Auctions; Marketplace Matching; Groups and Teams; Strategy; Education Industry; Sports Industry; United States

Educators

February 2022
Technical Note

Ethical Analysis: Fairness

By: Nien-hê Hsieh

Concerns about fairness can arise across a wide range of contexts. They include the treatment of others, how much things cost, how much workers are paid, the outcome of a decision, and how we assign benefits and burdens across individuals. What counts as fair in a... View Details

Keywords: Fairness; Management Practices and Processes; Decision Choices and Conditions

Educators

December 1996 (Revised June 1998)
Case

Midnight Networks, Inc.

By: H. Kent Bowen and Marilyn Matis

Midnight Networks, Inc., is a small computer network validation company. This case describes how the five founders built their business from operations earnings and how they established "best practices" operational processes to run their firm successfully. Operational... View Details

Keywords: Corporate Entrepreneurship; Business or Company Management; Operations; Organizational Culture; Applications and Software; Business Startups; Business Growth and Maturation; Information Technology Industry; Massachusetts

Educators

25 Jul 2017
First Look

First Look at New Research and Ideas: July 25, 2017

different types of seats are filled) on distributional objectives is comparable to the effect of adjusting reserve sizes. Either lowering the precedence order positions of reserve seats at a school or increasing the number of reserve seats weakly increases... View Details

Keywords: Carmen Nobel

November 2021
Article

Panel Experiments and Dynamic Causal Effects: A Finite Population Perspective

By: Iavor Bojinov, Ashesh Rambachan and Neil Shephard

In panel experiments, we randomly assign units to different interventions, measuring their outcomes, and repeating the procedure in several periods. Using the potential outcomes framework, we define finite population dynamic causal effects that capture the relative... View Details

Keywords: Panel Data; Dynamic Causal Effects; Potential Outcomes; Finite Population; Nonparametric; Mathematical Methods

2022
Working Paper

Statistical Inference for Heterogeneous Treatment Effects Discovered by Generic Machine Learning in Randomized Experiments

By: Kosuke Imai and Michael Lingzhi Li

Researchers are increasingly turning to machine learning (ML) algorithms to investigate causal heterogeneity in randomized experiments. Despite their promise, ML algorithms may fail to accurately ascertain heterogeneous treatment effects under practical settings with... View Details

Keywords: AI and Machine Learning; Mathematical Methods

Forthcoming
Article

Statistical Inference for Heterogeneous Treatment Effects Discovered by Generic Machine Learning in Randomized Experiments

By: Kosuke Imai and Michael Lingzhi Li

Researchers are increasingly turning to machine learning (ML) algorithms to investigate causal heterogeneity in randomized experiments. Despite their promise, ML algorithms may fail to accurately ascertain heterogeneous treatment effects under practical settings with... View Details

Keywords: AI and Machine Learning; Mathematical Methods

Find at Harvard

October 2007
Article

The Art of Designing Markets

By: Alvin E. Roth

Traditionally, markets have been viewed as simply the confluence of supply and demand. But to function properly, they must be able to attract a sufficient number of buyers and sellers, induce participants to make their preferences clear, and overcome congestion by... View Details

Keywords: Market Design; Market Participation; Market Transactions; Information Technology; Internet and the Web