Okun's Law Analysis with FRED Data

This project aims to empirically analyze Okun's Law, which suggests an inverse relationship between unemployment and GDP growth, using economic data from the Federal Reserve Economic Data (FRED). Data is collected, transformed, and analyzed to test Okun's Law across different U.S. states.

Workflow for Okun's Law analysis using FRED data

Project Structure

.
├── EXTRACT
│   ├── FREDDataFetcher.py         # Script to fetch data from FRED API
│   ├── gdp_data.py                # Script to extract GDP data from FRED
│   ├── state_codes.json           # JSON file mapping state names to codes
│   ├── unemployment_data.py       # Script to extract unemployment data from FRED
├── README.md                      # Project documentation
├── config.json                    # Configuration file for API keys and settings
├── database
│   ├── database_loader.py         # Script to load data into the database
├── dbt_fred
│   ├── dbt_project.yml            # dbt project configuration
│   ├── models                     # dbt models for data transformations
│   ├── tests                      # Tests for validation of transformations
├── sql
│   └── raw_tables                 # SQL scripts for raw data tables

Objective The goal of this project is to analyze how Okun's Law holds across various U.S. states. By using data on GDP growth and unemployment rates from FRED, we aim to quantify the relationship between economic growth and unemployment levels in line with Okun's Law.

Data Sources Gross Domestic Product (GDP) by State: Retrieved using gdp_data.py. Unemployment Rates by State: Retrieved using unemployment_data.py. State Codes: state_codes.json file maps state names to their respective codes used in FRED.

Set up config.json with your FRED API key and other configuration settings:

{
  "api_keys": {
    "fred_api_key": "e651cfc10fce0019d85cf23c192706e2"
  },
  "database": {
    "host": "localhost",
    "user": "postgres",
    "port": 5432,
    "password": "1234",
    "dbname": "fred"
  }
}

Create Tables and Schema

(in this example i use postgres but you can use a database of your choice)

• Make sure postgres is up and running
• create a new database named fed
• Run the file in ./sql/raw_tables/create_tables.py
• This creates the empty table structure for ingesting raw data into postgres Simply run

python create_tables.py

Data Extraction Fetching GDP Data:

Use gdp_data.py to extract GDP data by state:

python EXTRACT/gdp_data.py

Fetching Unemployment Data:

Use unemployment_data.py to extract unemployment rate data by state:

python EXTRACT/unemployment_data.py

The extracted data will be stored in a database or as CSV files based on the configuration.

Database Loading Load the extracted data into a PostgreSQL database using database_loader.py:

python database/database_loader.py

Data Transformation with dbt Run dbt to transform raw GDP and unemployment data into models ready for analysis:

cd dbt_fred
dbt run

Analysis

Once the data is processed, the analysis will focus on:

• Calculating the relationship between GDP growth and unemployment rates for each U.S. state.
• Visualizing trends to understand how closely the data aligns with Okun's Law.
• Cross-state comparisons to observe variations in how Okun's Law applies across different regions.

Grafana Dashboard output :

Incorporating other economic indicators (e.g., inflation) to see if they affect the relationship between unemployment and GDP growth. Extending the analysis to multiple time periods for longitudinal studies.