Published on Development Impact

Weekly links October 18: jumping from informal to formal, using AI in your teaching, noise is bad, and more…

This page in:
Young boxers at the White Collars Boxing Match 2019, taken by Mariajose Silva Vargas

·       “However, within-firm transitions from informal to formal jobs are not rare among formal firms’ employees. The quarterly predicted probability of transitioning from an informal to a formal job, without changing employers, is on average 14%.” – an amazing statistic from within Mexican firms in a VoxDev piece by Brenda Samaniego de la Parra and León Fernández Bujanda (based on a recent AEJ Applied paper). They also look at what happens when firms are randomly selected for worksite inspections on labor – there is a decline in formal hiring and increase in separations, but some of the currently informal workers transition to formal status. “However, by reducing the pool of informal workers, enforcement shuts down a critical channel for formal employment growth: within-firm informal-to-formal job transitions. This reduction in inflows, combined with a decline in new formal job creation, led to an overall decline in the formal workforce size at inspected firms.”

·       On the Econ that Matters blog, Vanisha Sharma uses a staggered DiD to look at the distributional impacts of 2G mobile phone roll-out in India, finding the richest quintile benefited a lot more than other quintiles.

·       YouTube video of Justin Wolfers discussing how econ educators should think about AI in the classroom. He argues that academics have been looking for ways ChatGPT is ineffective – whereas students have been looking for solutions and can solve most of the issues people think are there – so that the era of high-stakes at-home assessment is dead. He notes that students don’t know what they don’t know, so undirected chatbots aren’t great since students don’t know what to ask them. Undirected bots get used as a substitute for thinking, not as co-creators. So he argues for different ways of personalizing at scale: using as a Socratic tutor, as a practice exam coach, as a teaching collaborator (e.g. getting it to help with ideas for active learning), and summarizing unstructured student responses and feedback  -with nice examples of how these can be used.

·       Jacob Patterson-Stein has a new R package for flagging variables that potentially might contain personal identifying information.

·       On VoxDev, Joshua Dean summarizes his work on the productivity effects of noisy workplaces based on experiments with casual laborers in Nairobi doing sewing tasks and cognitive tests. “I find that when the engine is running outside of the room, the increased noise decreased productivity by 3%. For a firm, this change in productivity is equivalent to the impact of doubling the piece rate the participants were being paid for the pockets and sufficient to make investments in noise abatement cost effective….the same kind of noise decreases performance on the cognitive tests but does not decrease effort….I find both that willingness to pay for quiet is low and that it is not higher when receiving performance pay, suggesting workers are not aware of the effects of noise and will not act to mitigate their earnings losses.”

·       Conference call for papers: The NBER’s Working Group on Gender in the Economy will convene a research meeting on “Women in the Digital World” to explore the gender-specific effects of the digitization of economic and social activities. The digital economy, which encompasses developments from mobile money to electronic tracking of tax liability to online labor market platforms and the technology for work from home, has the potential to be transformative for the global population and to elevate the positions of women. The effects of access to cell phones, the internet, online marketplaces, and mobile money may differ by gender in ways that are mediated by and have the potential to shape women’s positions in their communities. Submissions of both empirical and theoretical research, of papers by scholars who are and are not NBER affiliates, and by researchers who are early in their careers or from historically under-represented groups in economics are welcome. Please do not submit papers that will be published by February 2025. To be considered for presentation, please upload submissions to https://conference.nber.org/confsubmit/backend/cfp?id=GEs25 by 11:59pm ET on Monday, December 9, 2024.

·       Reminders:

o    DECRG is hiring on the job market

o   Our call for blog your job market paper is now open.


David McKenzie

Lead Economist, Development Research Group, World Bank

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000