This repository contains a series of API calls that pull and aggregate labor market time series data from the public sources using julia programming language. The script builds a big data frame used in my co-authored research paper 'Uncertainty and Labor Market Fluctuations' by Jo and Lee (2019), Federal Reserve Bank of Dallas Working Paper. Other research papers citing our research is documented by the Google Scholar. The dataframe is best used to evaluate the U.S. labor market in the application of macroeconomics and machine learning.

Use of this repository requires a citation of this repository and a note of proper authorships of myself and the cited papers noted in the code.

The metrics calculated in the code use snippets translated from the MATLAB codes provided by the authors' research note below. All optimization processes uses julia's non-linear solver instead of MATLAB's fsolve.

For those with specific reasons, as a percent of total unemployed,

For those with unemployed less than 5 weeks, specifically,

The BLS series of unemployment by reasons are also broken down by different durations of unemployed.

Part-time employment count in non-agricultural sector, for,

With the collection of labor time series data, there are numerous applications to investigate the historical and real-time labor market and macroeconomic dynamics. Here is a growing list to explore in future research.

• State-space model implementation
• Recession predictor and space estimation of labor flow