Click to see the working paper

Crowdsourcing fundamentally transforms the way we work, particularly in creative industries or – on the other hand – in the execution of low-qualification tasks with platforms such as MTurk or oDesk. I’ve recently read a working paper about the latter, the marketplace for work oDesk (which has an army of researchers, mostly to analyze log data, see these cool visualizations). This paper particularly seeks to understand how culture impacts the attribution of work to people via oDesk. Or in other words: Do Indians from abroad attribute work more to Indians from the home country than to others, with similar qualification?

oDesk is a platform that connects workers who supply services with buyers who pay for these services. Examples include a variety of tasks that do not require intensive in-person interaction, such as data-entry or programming. The top skills currently listed on the oDesk platform are WordPress, SEO, translation, HTML, Facebook apps but also more creative tasks such as graphic design or article writing. I was stunned by the variety of services offered on the platform, similarly to websites like blur Group! Basically here’s how it works:

• A company or individual looking to hire fills out a description for each job, including the skills required, an assessment of the expected duration,
  and some preferred characteristics for a potential match
• Workers apply to the opening and bid an hourly rate
• After reviewing applications, the hiring contact at a firm sets up interviews
• The firm and selected worker agree to begin an employment relationship.

oDesk began operating in 2005, and is now “the world’s largest platform for online outsourcing,” according to the working paper. “There are other sites that provide similar services,” the authors say, “but oDesk’s expansion mainly reflects increasing demand for online labor services over time. […] The rate of overall growth for online outsourcing slowed with the financial crisis, but oDesk has continued to grow rapidly.” The working paper, called Diasporas and Outsourcing: Evidence from oDesk and India, has been authored by Ejaz Ghani (World Bank), William R. Kerr (Harvard Business School), and Christopher Stanton (University of Utah). As I said, they wanted to find out what role cultural ties (diaspora-based links) play in the attribution of tasks on oDesk.

According to LinkedIn, the name “Rajiv” is the most common Indian CEO name (image via http://www.pratham.name)

Since oDesk doesn’t collect peoples’ ethnicity or country of birth, they used names of company contacts and workers to probabilistically assign ethnicities (“the 4 most common surnames linked with the Indian ethnicity are Kumar, Singh, Ahmed, and Sharma,” they say). They analyzed data from 2005 to 2010 to find out whether Indians are more lilkely to choose Indian service providers over others. And indeed, they do: “Among the findings are that individuals with the surnames Gupta or Desai are more likely to be of Indian ethnicity than individuals with the surnames Ming or Hernandez,” they explain.

We find very clear evidence that diaspora-based links matter on oDesk, with ethnic Indians in other countries 32% (9 percentage points) more likely to choose a worker in India

Not only are ethnic Indians worldwide are more likely to contract with India when outsourcing, but this higher likelihood is even increasing over time! This indicates that on global marketplaces (similar to eBay and Amazon) dedicated to services (like oDesk or blur Group), culture does play a role in determining transactions. I’d like to highlight that, even though I use “crowdsourcing” in the title of the post, the term is not mentioned by the authors. They position their working paper as a piece to understand economic exchanges in general, not just crowdsourcing as a way to do business. “These findings contribute to understanding the role of diasporas and ethnic networks in economic exchanges across countries,” the authors say. It’s an interesting piece, comparable to similar papers about MTurk, and that might trigger more research in that area. Looking forward to it!

Any comments or remarks? Let me know