This project investigates the determinants of Finger Millet (Ragi) production across Indian districts using modern econometric techniques. By combining agricultural, environmental, and input data, we estimate and interpret both Cobb-Douglas and quadratic production functions. The goal is to understand how land, fertilizer, rainfall, and soil conditions affect output, and to provide evidence on input complementarities and returns to scale.

Finger Millet is a climate-resilient crop important for food security in semi-arid regions. Understanding the drivers of its productivity can inform policy, extension services, and farmer decision-making, especially under changing climate and resource constraints.

• ECO221_Project_2025_Final.csv: Main raw dataset with district-level variables (production, area, fertilizer use, etc.).
• RF_DistrictWise_ECO221_2025.csv: District-wise rainfall data.
• Salinity_Alkalinity_ECO221_2025.csv: Soil salinity/alkalinity data.
• Finger_Millet_Data.csv: Cleaned and merged data.
• Finger_Millet_Prepared_for_Cobb_Douglas.csv: Data prepared for Cobb-Douglas regression.
• Finger_Millet_Prepared_for_Quadratic.csv: Data prepared for quadratic regression.
• Finger_Millet_with_Rainfall_and_Soil_Cleaned.csv: Data with rainfall and soil variables cleaned.
• Inconsistent_Rows.csv: Rows with missing or inconsistent data.
• Presentation.pptx: Final project presentation slides.
• Project_ECO221_2025.pdf: Project guidelines and report.

1. Data Cleaning & Preparation
   • Handle missing values and zeroes in key variables (e.g., irrigated area, production).
   • Create new variables for analysis (e.g., irrigatedarea_new, unirrigatedarea).
   • Merge rainfall and soil data.
   • Remove or impute inconsistent or missing data as documented.
2. Modeling
   • Estimate Cobb-Douglas and quadratic production functions using OLS regression.
   • Test for returns to scale and input complementarity (e.g., irrigation-fertilizer interactions).
   • Interpret coefficients, elasticities, and marginal effects.
3. Diagnostics & Interpretation
   • Check for multicollinearity, non-linearity, and residual issues.
   • Discuss economic implications, limitations, and robustness.
4. Reporting
   • Summarize findings in the notebook, presentation, and final report.

• code.ipynb: Main Jupyter notebook with all code, analysis, and interpretation.
• Data files as listed above.
• Presentation.pptx: Slide deck summarizing the project.
• Project_ECO221_2025.pdf: Final report and guidelines.

1. Open code.ipynb in Jupyter Notebook or VS Code.
2. Run the notebook cells sequentially to reproduce the analysis.
3. Review the outputs, plots, and interpretations provided in the notebook.
4. Refer to the presentation and report for summarized findings and recommendations.

• Python 3.x
• pandas
• numpy
• statsmodels
• matplotlib
• seaborn
• (Optional) Jupyter Notebook or VS Code for interactive analysis

Install dependencies with:

pip install pandas numpy statsmodels matplotlib seaborn

• Returns to Scale: The Cobb-Douglas model estimates the sum of elasticities to assess returns to scale.
• Input Complementarity: Quadratic and interaction terms test whether fertilizer effectiveness increases with irrigation.
• Policy Implications: Results can inform recommendations on optimal input use and highlight the importance of environmental factors.

• The analysis is limited by data quality, missing values, and potential measurement errors.
• Results are specific to the 2017 dataset and may not generalize to other years or regions.
• Multicollinearity and omitted variable bias are addressed but may still affect some estimates.

• Sanyam Barwar
• Pardeep Singh
• Anusha Anand
• Amarty Singh

This project is for academic use in the ECO221 Econometrics course at IIIT Delhi.

For questions or collaboration, please contact the authors via the IIIT Delhi student portal or email.

1. Data Cleaning & Preparation
   • Handle missing values and zeroes in key variables (e.g., irrigated area, production).
   • Create new variables for analysis (e.g., irrigatedarea_new, unirrigatedarea).
   • Merge rainfall and soil data.
2. Modeling
   • Estimate Cobb-Douglas and quadratic production functions using OLS regression.
   • Test for returns to scale and input complementarity.
   • Interpret coefficients and model fit.
3. Diagnostics & Interpretation
   • Check for multicollinearity, non-linearity, and residual issues.
   • Discuss economic implications and limitations.
4. Reporting
   • Summarize findings in the notebook, presentation, and final report.

1. Open code.ipynb in Jupyter Notebook or VS Code.
2. Run the notebook cells sequentially to reproduce the analysis.
3. Review the outputs, plots, and interpretations provided in the notebook.

• Python 3.x
• pandas
• numpy
• statsmodels
• matplotlib
• seaborn
• (Optional) Jupyter Notebook or VS Code for interactive analysis

Install dependencies with:

pip install pandas numpy statsmodels matplotlib seaborn

• Sanyam Barwar
• Pardeep Singh
• Anusha Anand
• Amarty Singh

This project is for academic use in the ECO221 Econometrics course at IIIT Delhi.