This provides total job loss estimates by industry for every 2018 Census CBSA in the US (excluding Puerto Rico, Guam, and other territories) due to Covid-19. Couple of notes about the data:

1. The attributes in the file starting with X correspond to LEHD industry codes, and the translation list can be found in industry_code_translation_list.csv.
2. These estimates for jobs loss are based on place of residence for workers. So when we report that 100 jobs are lost in CBSA j, this means that there are 100 workers who live in CBSA j that have lost jobs.
3. The columns labelled max and tmax are the max values by tract and industry within counties/cbsas and only used for data visualization. Most users can safely ignore these columns as they were for our internal use in the data viz
4. Every row in the CSV is a unique census CBSA.

Below is a description of each of the columns in the data:

• cbsa: The 5 digit FIPS code for the CBSA
• cbsa_name: Human readable CBSA name
• state_name: Full state name
• X01 - X20: Number of estimated jobs lost in that particular industry by residents of that particular CBSA. The translation from X01 to human readable industries can be found in industry_code_translation_list.csv.
• X000: The total number of jobs lost by residents of that particular CBSA
• max and tmax: bucketed max values by tract and industry, can be safely ignored
• total_workers_employed: Total number of workers who live in that CBSA
• worker_job_loss_rate: The percentage of all jobs that are lost, or X000 / total_workers_employed. The numerator is our estimated jobs lost and the denominator is the total number of jobs as estimated by the Census LODES data. Note this number may be slightly off due to rounding