Technology Adoption and Firm Employment: Evidence from Burundi
How to cite
Naso, Pedro; Ndayikeza, Michel Armel. 2024. "Technology Adoption and Firm Employment: Evidence from Burundi." AEA RCT Registry. AEARCTR-0013134. https://doi.org/10.1257/rct.13134-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Introducing more advanced technologies into firms holds the promise of elevating job quality and productivity, fostering higher wages and more efficient societal outputs. However, in developing countries, data on how technology adoption influences factors such as employment rates and wages remain sparse. To bridge this gap, our study focuses on the textile sector in Burundi, where most small firms rely solely on manual
sewing machines. Through a randomized controlled trial (RCT), we will equip selected firms with automatic sewing machines and offer operation training. Our analysis will track short-term (3 months) and long-term (9 months) effects on three key variables: employment, productivity, and real wages.
Intervention
We will equip selected dress-making firms in Burundi with automatic sewing machines and offer operation training.
Experimental design
We assume that technology affects labour via three channels: (i) replacement, (ii) reinstatement, and (iii) real income effect (Hötte and Theodorakopoulos (2022)). Replacement refers to technology-induced job losses. Reinstatement refers to technology-induced job creation. Real income effect refers to new technology raising real wages via an increased marginal product of labour. The historical account demonstrates that, at the aggregate level, channels (ii) and (iii) overcome channel (i). However, replacement can become a strong force at the micro level in the short run. We will run an RCT with textile firms in Burundi to measure the relative importance of each of these channels. We will have a sample of small dressmaking firms that use manual sewing machines to produce dresses and other relatively expensive clothes. These firms are usually composed of around 3 employees and are prevalent both in the main cities,
such as Bujumbura and Gitega, and in small towns in the countryside. We will randomly lend to a subset of these firms automatic sewing machines and power banks at a low cost.
Automatic sewing machines, usually preferred by professional tailors, are faster and more efficient than manual sewing machines. We will also provide training to tailors in treated firms on how to operate these machines.
We will run three main surveys with our sample of firms. The first survey will happen before the intervention and will collect baseline data on firms’ characteristics, such as number of employees, wages, sales and revenue, and information on firm history. The second survey
will happen 3 months after the intervention and will measure the short run effects of technological adoption. The final survey will happen after 9 months of the intervention and will collect the same variables of the second survey.
Design details
Cited by
No citation data available for this registration yet.