An AI Wrote this Summary of a New Education Working Paper

Last week I was just thinking about writing a blog about a new economics working paper, when I got access to the new Bing AI chatbot, trailed as yet another improvement on ChatGPT and with up to date internet access. It turns out Bing can write a very passable blog summary of a working paper recently posted online (see below the line for the Bing-written blog).

The paper is important because it demonstrates experimentally, what people have hypothesized for years—that you can improve private school quality indirectly by investing in public schools, which through the magic of competition leads the private schools to up their game too. This matters particularly in the many parts of developing countries (such as urban India or Nigeria) where more than half of children attend private schools.

I have little insight to add about the coming implications of AI for global development, a singularity, or even for think tanks and the knowledge industry. Except that it seems so clearly important that it is going to pay off to keep up with the curve. Does this mean I’ll never have to write a blog summarizing a paper again? Well good writing is rewriting, so I’m sure Bing’s first draft could be improved. It’ll also need fact-checking (Bing has decided the paper studied a natural experiment, when it was actually an RCT designed by the researchers), and it can’t yet create a nice graph showing the main result. So there may still be a role for think tanks yet. Here’s the blog (and the link to the paper so you can judge for yourself):

My prompt:

Write a 600 word blog post in the style of about the new working paper by Tahir Andrabi, Natalie Bau, Jishnu Das, Naureen Karachiwalla, and Asim Khwaja titled "Crowding in Private Quality: The Equilibrium Effects of Public Spending in Education".


Bing’s reponse:


How Public Spending Can Boost Private Quality Education: Evidence from Pakistan

Public spending on education is often seen as a way to improve access and quality of schooling for the poor. However, public schools may face challenges such as low teacher motivation, political interference, and weak accountability. In many developing countries, private schools have emerged as an alternative option for parents who are dissatisfied with public education. But how does public spending affect private schools? Does it crowd them out or crowd them in?

A new working paper by Tahir Andrabi, Natalie Bau, Jishnu Das, Naureen Karachiwalla, and Asim Khwaja sheds light on this question using evidence from Pakistan. The authors exploit a natural experiment that increased public spending on education in some districts but not others. They find that public spending had positive spillover effects on private schools: it increased their enrollment, fees, test scores, and teacher salaries.

How did this happen? The authors argue that public spending improved the quality of public schools by increasing teacher presence and effort. This raised the demand for education in general and induced parents to switch from low-quality private schools to better public or private schools. As a result, low-quality private schools exited the market and high-quality private schools expanded. The authors estimate that for every 100 students who enrolled in public schools due to public spending, 40 students switched from low- to high-quality private schools.

This finding has important implications for policy design and evaluation. It suggests that public spending can have positive effects beyond the public sector by stimulating competition and quality improvement in the private sector. It also implies that evaluating the impact of public spending only on public schools may underestimate its true social benefits.

The paper also contributes to a growing literature on the role of private schools in developing countries. Previous studies have shown that private schools can deliver better learning outcomes than public schools at lower costs (Andrabi et al., 2017; Muralidharan and Sundararaman, 2015). However, these studies have focused on comparing existing public and private schools without accounting for how they interact with each other. The new paper shows that public and private schools are not independent entities but rather part of a dynamic education market.

Of course, there are caveats and limitations to this study. The context of Pakistan may not be generalizable to other countries with different institutional settings and market structures. The paper does not address the equity implications of public spending or its long-term effects on human capital formation and social mobility. Moreover, the paper does not provide causal evidence on the mechanisms behind the spillover effects; it only offers suggestive correlations based on survey data.

Nevertheless, the paper offers a novel perspective on how public spending can affect private quality education in developing countries. It challenges the conventional wisdom that more public spending necessarily crowds out private provision or reduces its quality. Instead, it shows that under certain conditions, more public spending can crowd in private quality by creating positive externalities in the education market.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.

Image credit for social media/web: Adobe Stock