Data Research Associate, Using Machine Learning for improving GST Collection in Punjab - J-PAL South Asia

J-PAL Office 
Language Requirement 
Length of Commitment 
One year
Start Date 

J-PAL South Asia at the Institute for Financial Management and Research seeks qualified applicants for the position of Data Research Associate (RA) for the “Using Machine Learning for improving GST Collection” project in Punjab. The position offers an opportunity to gain first-hand data science as well as field management experience in an organization undertaking cutting-edge development research.

The RA, who will be based out of Mohali, will report to the Principal Investigators (PIs) from University of California, Berkeley and work under their direct supervision. An ideal candidate is comfortable with quantitative analysis, has had extensive experience working with data, and is passionate about engaging closely with government to apply data science and adapt to the challenging environment of public policy for social good. The position entails minimal in-the country travel.

Project Description

Improving the state’s ability to tax effectively is increasingly seen as central to the development process. The launch of Goods and Service Tax (GST) was seen by many observers as a key milestone towards this goal. However, whether this goal is achieved remains to be seen.

This collaboration between the Punjab Excise and Taxation Tax Department (ETD) and the Abdul Latif Jameel Poverty Action Lab South Asia at IFMR (JPAL-SA) proposes a multi-pronged approach to tackle compliance problems faced by the ETD. A significant component of the project proposes using “big-data” methods to reduce the burden on tax officials in identifying tax evasion. Tax officials currently are able to identify tax evading firms only after manually evaluating available information. JPAL-SA intends to complement this approach by using “big-data” methods to help in two ways: reducing the effort needed to find such firms and finding more such firms. JPAL-SA then also intends to complement the “big-data” approach with other evidence-based methods.

Key Responsibilities

The RA will work closely with the academic researchers and the government officials on a variety of tasks including, but not limited to the following:

  • With close coordination and support from the PIs, the RA will carry out data cleaning and comprehensive data analysis using Python, machine learning modules (H2o, graphlab), and Stata.
  • The RA will need to work from the ETD office (Mohali) to replicate the coding and analysis provided remotely by the PIs and debugging all unexpected errors while ensuring data privacy and security protocols are met at all times
  • The RA will proactively liaise and communicate with government stakeholders to ensure alignment on key decisions around the research design and project activities.
  • The RA may need to interview and/or work closely with experts such as Chartered Accountants/department officials from the ETD to understand the GST enforcement system and communicate the same to the PIs
  • The RA will assist in preparation of documents and presentations for donor reporting as well as dissemination activities.
  • The RA will ensure all meetings and discussions are formally documented.
  • The RA may need to assist in designing survey questionnaires, coding the questionnaires for digital data collection, conducting qualitative research, and testing the instruments.
  • The RA will carry out any other task as assigned by the PIs.

Desired Qualifications/Attributes                    

  • A Bachelor's/Master's degree that preferably includes courses in computer programming, data science, machine learning, economics, statistics, social sciences, public policy, or related fields.
  • The RA should be either a social scientist who is comfortable with data-related coding, or a programmer willing to learn and work on evidence based analysis.
  • Experience with Python is preferred
  • Ability to create neat, replicable, documented code. Ability to sanity-check code, resolve bugs and handle problems independently.
  • Ability and motivation to quickly learn new technologies or platforms that may be required for research or execution (e.g. Python, Stata, PostgreSQL, SurveyCTO/ODK)
  • Strong quantitative skills.
  • Desire and ability to become familiar with randomized control trials.
  • Excellent management and organizational skills.
  • Ability to manage crucial relationships with government officials, and ensuring smooth coordination.
  • Attention to detail and advanced communication skills (both written and verbal) for interaction with relevant stakeholders.
  • Flexible, self-motivating, able to multitask and work without much supervision.
  • Ability to deal with unstructured and changing environments.
  • Ability to foresee challenges, and communicating them before they happen.

How to Apply

Complete the J-PAL/IPA common application. After completing a brief registration, choose the [Research] job category, then select "[Data Research Associate, Using Machine Learning for improving GST Collection in Punjab - J-PAL South Asia]” and include resume, cover letter, transcripts, and two recommendation letters. Note that applications with incomplete submissions will not be considered.

You are able to submit your application before your references complete their recommendation letters; however, you must include their contact information before submitting your application. Given the volume of applications received, only short-listed candidates will be contacted for an interview.