Matthew Olckers

What Type of Explanation Do Rejected Job Applicants Want? Implications for Explainable AI

with Alicia Vidler and Toby WalshRejected job applicants seldom receive explanations from employers. Techniques from Explainable AI (XAI) could provide explanations at scale. Although XAI researchers have developed many different types of explanations, we know little about the type of explanations job applicants want. We use a survey of recent job applicants to fill this gap. Our survey generates three main insights. First, the current norm of, at most, generic feedback frustrates applicants. Second, applicants feel the employer has an obligation to provide an explanation. Third, job applicants want to know why they were unsuccessful and how to improve.

Incentives to Offer Algorithmic Recourse

with Toby WalshDue to the importance of artificial intelligence (AI) in a variety of high-stakes decisions, such as loan approval, job hiring, and criminal bail, researchers in Explainable AI (XAI) have developed algorithms to provide users with recourse for an unfavorable outcome. We analyze the incentives for a decision-maker to offer recourse to a set of applicants. Does the decision-maker have the incentive to offer recourse to all rejected applicants? We show that the decision-maker only offers recourse to all applicants in extreme cases, such as when the recourse process is impossible to manipulate. Some applicants may be worse off when the decision-maker can offer recourse.

Gamblers Learn from Experience

with Joshua BlumenstockMobile phone-based sports betting has exploded in popularity in many African countries. Commentators worry that low-ability gamblers will not learn from experience, and may rely on debt to gamble. Using data on financial transactions for over 50 000 Kenyan smartphone users, we find that gamblers do learn from experience. Gamblers are less likely to bet following poor results and more likely to bet following good results. The reaction to positive and negative feedback is of equal magnitude and is consistent with a model of Bayesian updating. Using an instrumental variables strategy, we find no evidence that increased gambling leads to increased debt.

Predicting Attendance in a Health Program to Prevent Mother-To-Child Transmission of HIV: Comparing Community Health Worker and Machine Predictions

with Alexander Lam, Lazola Makhupula, and Mhlasakululeka Mvubu

Accelerating the Adoption of Mobile Money

with Joshua Blumenstcok and Suraj R. Nair