DiSP:  Dispersion Statistics on Productivity - Retail

September 2025

The Census Bureau has reviewed this data product to ensure appropriate access, use, and disclosure avoidance protection of the confidential source data. (Project No. P-7526913, Disclosure Review Board (DRB) approval number: CBDRB-FY25-0421).	

CSV files:
output_per_hour_estab.csv contains dispersion statistics for establishments (stores)							
output_per_hour_firm.csv contains dispersion statistics for firms	

Variable names in csv files:
naics4 - 4-digit industry identifier
year   - time identifier
sd     - standard deviation
d7525  - difference between the 75th and 25th percentile of the within-industry distribution of the natural log of productivity in a given year
d9010  - difference between the 90th and 10th percentile of the within-industry distribution of the natural log of productivity in a given year
d1001  - difference between the 10th and 1st percentile of the within-industry distribution of the natural log of productivity in a given year
d9990  - difference between the 99th and 90th percentile of the within-industry distribution of the natural log of productivity in a given year	
d9075 – difference between the 90th and 75th percentile of the within-industry distribution of the natural log of productivity in a given year
d7550 – difference between the 75th and 50th percentile of the within-industry distribution of the natural log of productivity in a given year
d5025 – difference between the 50th and 25th percentile of the within-industry distribution of the natural log of productivity in a given year
d2510 – difference between the 25th and 10th percentile of the within-industry distribution of the natural log of productivity in a given year
fraction_variance_within_firm – fraction of within-industry variation coming from within-firm variation [variable in output_per_hour_firm.csv only]
hours_share – industry share of total hours worked
*      - activity weighted dispersion, where the weight is defined by the denominator of the productivity indicator (total hours)
								
Note: Cells with identified confidentiality concerns are suppressed with a publication flag ‘D’. The interdecile range (90-10) is calculated after disclosure avoidance review as the sum of the following ranges: 90-75, 75-50, 50-25, and 25-10. Therefore, the interdecile range will contain a ‘D’ if one or more of its' components contains a ‘D’.