IGL Trials Database

We examine the productivity effects of a generative artificial intelligence technology—the assistive chatbot ChatGPT—in the context of mid-level professional writing tasks. In a preregistered online experiment, we assign occupation-specific, incentivized writing tasks to 444 college-educated professionals, and randomly expose half of them to ChatGPT. Our results show that ChatGPT substantially raises average productivity: time taken decreases by 0.8 SDs and output quality rises by 0.4 SDs.

Entrepreneurs in developing countries face a series of diverse constraints to growth, including lack of access to business skills, markets, and finance. The binding constraints vary from firm to firm, implying that the returns to possible interventions are likely to be heterogeneous.

We evaluate the impact of a training program aimed at improving the relational atmosphere in the workplace. The program encourages prosocial behavior and the use of professional language, focusing primarily on leaders’ behavior and leader-subordinate interactions. We implement this program using a clustered randomized design involving over 3,000 headquarters employees of 20 large corporations in Turkey. We evaluate the program with respect to employee separation, pro- and antisocial behavior, the prevalence of support networks, and perceived workplace climate.

We show in a large-scale field experiment that a brief exposure to female role models working in scientific fields affects high school students’ perceptions and choices of undergraduate major. The classroom interventions reduced the prevalence of stereotypical views on jobs in science and gender differences in abilities. They also made high-achieving girls in grade 12 more likely to enrol in selective and male-dominated science, technology, engineering and mathematics programs in college.

We document interest in labor reallocation among small firm owners in Ghana; 60% and 41%, respectively, self-report willingness to hire or work for the average local firm owner. Firm owners also exhibit high willingness-to-pay for information on a random subset of hiring firms and jobseeking firm owners, during a Becker-Degroot-Marschak exercise. Conditionally random variation in access to this information generates immediate labor adjustments within and between firms, though rarely of firm owners themselves, and impacts firm closure 5-months post-intervention.

Can a set of low-cost behavioural nudges encourage more small businesses to adopt productivity-raising digital technologies? This randomised controlled trial sought to test whether businesses could be nudged into using a cloud-based system to improve the efficiency of invoice processing. All participants in the trial were offered access to the system free of charge for a 12-month period, with a treatment group receiving weekly email reminders to make use of the system.

Interference across competing firms in RCTs can be informative about market structure. An experiment that subsidizes a random subset of traders who buy cocoa from farmers in Sierra Leone illustrates this idea. Interpreting treatment-control differences in prices and quantities purchased from farmers through a model of Cournot competition reveals differentiation between traders is low. Combining this result with quasi-experimental variation in world prices shows that the number of traders competing is 50 percent higher than the number operating in a village.

Across developing economies, cash is the conduit for retail transactions. Policymakers, multinational product manufacturers and marketers of electronic payment systems are interested in understanding how to stimulate the growth of electronic payments in emerging markets. In this paper, we investigate what hinders the adoption of e-payment technology by traditional retailers, in particular, whether barriers to adoption are technological, informational or financial in nature.

Differences in management quality are an important contributor to productivity differences across countries. A key question is then how to best improve poor management in developing countries. We test two different approaches to improving management in Colombian auto parts firms. The first uses intensive and expensive one-on-one consulting, while the second draws on agricultural extension approaches to provide consulting to small groups of firms at approximately one-third of the cost of the individual approach.

We examine the influence of physical proximity on between-startup knowledge spillovers at one of the largest technology co-working hubs in the United States. Relying on the random assignment of office space to the hub's 251 startups, we find that proximity positively influences knowledge spillovers as proxied by the likelihood of adopting an upstream web technology already used by a peer startup.

A randomized control trial with 945 entrepreneurs in Jamaica shows positive shortterm impacts of soft-skills training on business outcomes. The effects are concentrated among men, and disappear twelve months after the training.

This experiment tries to understand how managers respond to uncertainty when making research and development decisions. Three experiments were conducted with master’s degree students in a program focused on the intersection of business and technology.

What is preventing entrepreneurs and managers from forming peer connections themselves? This paper argues that entrepreneurs may be under-networked because they lack the necessary social skills that allow them to match efficiently with knowledgeable peers.

The limited market size of many small emerging economies is a key constraint to the growth of innovative small and medium enterprises. Exporting offers a potential solution, but firms may struggle to locate and appeal to foreign buyers. A six-country randomized experiment was conducted with 225 firms in the Western Balkans to test the effectiveness of 30 hours of live group-based training and 5 hours of one-on-one remote consulting in overcoming these constraints.