README FILE
CENSUS 2000 SPECIAL TABULATION PRODUCT 86 (STP86) -- ASCII VERSION
COUNTY OF RESIDENCE BY COUNTY OF WORK BY BUREAU OF ECONOMIC
ANALYSIS (BEA) INDUSTRIAL CODE STRUCTURE BY MEAN EARNINGS
INTRODUCTION
------------
This Census 2000 special tabulation is being distributed on CD-ROM disc. The
CD contains four files, as described below.
1) stp86-ascii.txt - the file you are reading now. A text file containing
the documentation for the product.
2) cty2cty.dat - the ASCII data file, as described below.
3) BEA-codes-11.xls - a Microsoft Excel document that shows how each of
the BEA categories used in the product were defined.
4) st_cnty.txt - a text file containing the FIPS state and county codes,
county name, and state postal abbreviation for each county of residence.
Note that the codes are slightly different for place of work geography
(see the Data Description section for the WSTATE and WCOUNTY variables).
FILE DESCRIPTION
----------------
The ASCII version of STP86 consists of one data file based on the Census
2000 long form (sample) data. Each record in the file represents a unique
combination of a particular residence county, a particular workplace county,
and a particular BEA industry code. The county-of-residence by county-of-work
commuter flow file uses 11 BEA categories. The file shows the weighted
worker estimate for flows using a threshold or minimum size of 3 unweighted
worker records. Only county-to-county flows that are comprised of at least 3
unweighted worker records are shown in this file. All other flows have been
omitted from the file. Within the qualifying flows, mean earnings figures
are only presented for BEA categories for which there were three or more
unweighted worker records from which to calculate the mean. Cells with too
few records to calculate the mean have an upper case D in parentheses, (D),
as the value in the mean earnings cell. The filename for this dataset is
cty2cty.dat and there are 335,387 records in the file.
DATA DESCRIPTION
----------------
Variable Length Location Description
------------- ------ -------- -------------------------------------------
RSTATE 2 1 - 2 FIPS State Code of Residence
01-56 AL through WY state code
RCOUNTY 3 4 - 6 FIPS County Code of Residence
001-840 County code
WSTATE 3 8 - 10 State Code of Workplace (modified FIPS)
001-056 AL through WY (with leading zero)
072 Worked in Puerto Rico
301 Worked in Canada
303 Worked in Mexico
556 Worked in other foreign country
557 Worked in US Island Area or
elsewhere
WCOUNTY 3 12 - 14 FIPS County Code of Workplace
001-840 County code
999 Worked outside 50 states, DC,
and Puerto Rico
IND11 2 17 - 18 BEA Industrial Code
01 Agric., forestry, fishing, mining
02 Utilities, transportation and
warehousing
03 Construction
04 Manufacturing
05 Wholesale trade and retail trade
06 Information, finance, insurance,
real estate, and rental/leasing
07 Services (except private households)
08 Private households
09 Federal civilian and military
10 State and local government
11 Self-employed (part) and unpaid family
workers
MEAN_EARNINGS 8 19 - 26 Mean Earnings in 1999 of Workers in Flow
(D) Less than three cases upon which to
calculate mean earnings
WORKERS 9 27 - 34 Number of Workers in the Flow in the
Specified BEA Category
BEA INDUSTRIAL CODE STRUCTURE
-----------------------------
See the companion file, BEA-codes-11.xls, for information on the 11-category
BEA industrial structure used in this product. For each category the Census 2000
class of worker codes and industry codes used to define the category are shown,
as well as a title for the group.
TECHNICAL NOTES
---------------
1) Rounding - The estimates of workers in this special tabulation have been
rounded. Values from 1 through 7 have been rounded to 4. Values of 8 or greater
have been rounded to the nearest multiple of 5, unless the estimate already
ended in 5 or 0, in which case it was not changed. As a result, estimates
derived from this file may not be identical to comparable figures contained
in other census products. The greater the number of records from this file
that are summed for comparison purposes, the more rounding error there may
be and the greater the difference between the estimates from different sources
may be.
2) Workers - In this special tabulation, workers are defined as people 16 years
and older who were employed and at work during the Census reference week. This
is the week prior to when the questionnaire was filled out, for most people the
last week of March 2000. Workers include both civilians and people in the Armed
Forces, and part-time workers as well as full-time. The concept of a worker
excludes people who had a job but did not work during the reference week.
Examples of people in the excluded category would be those who didn’t work
during the reference week because they were sick, on vacation, laid off, or
otherwise absent from work for the week. Nationwide, about 2 percent of
employed persons fell into this category.
3) Earnings in 1999 - The earnings data collected in Census 2000 are amounts
for calendar year 1999. For workers whose class of worker code was six
(self-employed, not incorporated), their self-employment income was used as
their earnings. For workers whose class of worker code was seven (self-employed,
incorporated) or eight (unpaid family workers), their self-employment income
and their wage and salary income were combined and that amount was used as
their earnings. For all other workers, their wage and salary income was used
as their earnings. Since some of the earnings amounts reflect self-employment
income, it is possible that certain mean earnings amounts in the file will be
less than zero (i.e., negative dollar amounts).
Not every person who was working at the time of Census 2000 had earnings
during 1999. The estimate of workers in each industry category in the county-to-county flows includes both workers with earnings in 1999 and those without
earnings in 1999. However, the calculation of mean earnings is based only upon
workers who actually had earnings in 1999. To calculate mean earnings, the
aggregate earnings amount for workers with earnings in 1999 was rounded to
the nearest hundred dollars (-$99 to $99 were rounded to zero). Then the
rounded aggregate earnings amount was divided by the estimate of workers in
that industry category in the flow who reported earnings in 1999. The result
was rounded to the nearest dollar for presentation in the files.
FURTHER INFORMATION
-------------------
Questions or requests for further information about Census 2000 Special
Tabulation Product 86 should be referred to the Journey-to-Work and Migration
Statistics Branch of the U.S. Census Bureau at (301) 763-2454.