  |
 |
 |
 |
NRIAI Home
About Us
News
Technical Resources
Partnerships
Contact Us
|
   |
|
|
Documentation for the PRMS Potential Clients databaseUses for this dataThe primary purpose for this data is for populating the Potential Clients and Potential Ag. Clients database in PRMS Additional uses are for state EEO coordinators and for EQIP coordinators. The Zip-code data is provided for demographic breakdowns at a finer level for some field office areas, and for watershed and EQIP projects. Please contact David Buland (David.Buland@usda.gov) | Tel: 254-770-6522 | for additional information. BackgroundThe database tables in FOCS at each field office has fields for Potential
Agricultural and Non-Agricultural Clients by race and sex for the field office
service delivery area. In most many field offices this area is defined by the
county boundaries. We are also providing data at the zip-code level for
demographic breakdowns at a finer level for some field office areas, and for
watershed and other project related actives if needed. Broad guidelines were
given on filling out these fields in the past, however, these guidelines were
considered to be too broad to be useful for national analysis. As PARS progress
reporting expanded, and civil rights parity information increased in
significance to national planning, the shortcomings of the existing database
figures became apparent. In August, 1997, a group headed by the Operations
Management and Oversight Division (OMOD), and the Civil Rights and Program
Compliance Division (CRCRP) defined a national standard for this database, and
started the procedure to develop the base data. Because of existing software,
the group was locked into the racial categories: White(W), Black(B), Native
American(I), Asian(A), and Hispanic(H). Hispanic was treated as a race, not an
ethic group, in the NIMS reporting as indicated in both the Census and Ag.
Census data. Both base data sources needed to be modified to delineate Hispanics
from the five existing ('Other' was the fifth) groups for FOCS/NIMS. This
mis-use of Hispanic Origin as a race was corrected in the PRMS software, and the
current databases. The older FOCS databases are still available for any field
offices still using them.
Calculation StepsCalculations steps for the Agricultural Clients:NASS delivered an 400 MB ASCII dataset for the 1,649,272 operators (farms) replying to the 1992 Census with state, county, FIPS, zip code, weight (1 or 2), and race/sex/Hispanic origin of the operator. The only complication was that the 'Other' category was omitted, but could be estimated from the 8,229 observations missing a racial category. These are summarized as others. The weights of 1 or 2 were used to adjust up to the total 1,925,300 farms by counting specific representative farms twice.Hawaii is the only state with recorded 'Others'; 4 Hispanic, and 1,939 non-Hispanic Other. In Hawaii, the survey asked for five subgroups, Hawaiian, Japanese, Chinese, Korean, and Filipino, but did not ask for other Asian. In the current dataset, all Hawaii subgroups are summarized as Asian. The local NASS office can provide a better count. The decision was made at the workshop last summer to use Ag. Census data even recognizing the survey bias against counting female operators. There is no other reliable data source for agricultural operators. The final counts will include the five PRMS racial categories ( White, Black, Native American, Asian, and Other (Missing)); by Male and Female, and by Spanish/non-Spanish origin. . This count matches the total numbers of farms. The county/zip code table has to remain in a database or text format due to size limitation. Calculations steps for the Non-Agricultural Clients:We are using Owner-Occupied housing by race, and Hispanic origin, allocated by gender using general population ratio by racial group. The data source is CensusCD from Gelytics, Inc. This calculation was first done for counties, and then repeated for zip codes. There is no method using this data source to obtain County/Zip code combinations. If this procedure is redone after the 2000 Census, USDA should use the base data source as done with the Ag. Census data.All racial groups by age by gender were moved out from CensusCD to dbf files and then to the Excel spreadsheets: OwnerOccupied.xls, WBNAagegender.xls, Hispanic.xls, and AOHagegender.xls. For some unknown reason, the dbf files did not load directly into Access. We added six columns into the spreadsheets; WBNAagegender.xls and AOHagegender.xls and summed the total of all age group by age and gender for the 3,141 counties. These total columns were then put into Clients.xls, sheet POPraceGender for total county population by race and gender. Checking the totals of all counties with the total US population shows a missing 494,743 people, out of 248,709,873, or 0.2%. Each race was missing less than 0.5%, except for Other, which had an exact count of 9,710,156. I assume that these are people for whom Race was known, but not age or county, so they were not included in the county/age/race/gender counts. Since we are only using the % gender in these calculations, being within 0.2% was good enough to proceed. See the printouts for the totals. There are currently 270 million people (mid 1998), a 8.5% increase since the 1990 census. While this is an acceptable national increase, some high-growth local areas may need to update their population estimates for FOCS/NIMS. 1995 total population figures for each county are available at http://www.census.gov/datamap/www/index.html Nationwide, 90% of the Other Race is Hispanic. The remaining Others (±10%) will be classified as White. Hawaii may want to adjust that calculation, placing their 'other' into Asian/Pacific Islander. I determine the general population estimates by White, Black, Native American, Asian, and Hispanic by Gender by multiplying the % Hispanic by Race/Gender, and adding the remaining Other category to White. This provides the general population estimates by NRCS categories of Race/Gender. Most counties have an exact match by category to total population. The largest rounding error is LA County, with an extra 500 people out of 9 million. Then the percent Male was calculated for each category in AdjRaceSex. The spreadsheet OwnerOccupied.xls has Owner Occupied by Race, and Hispanic Owner Occupied by Race. Hispanic Owner Occupied by Race was totaled to get total Hispanic Owner-Occupied, and Adjusted non-Hispanic Owner-Occupied by subtracting the Hispanic Owner Occupied by race. Others were added to Whites. These Adjusted Owner-Occupied columns were copied to the worksheet OwnerOccupied in Clients.xls. Then these were multiplied by the percent male from step 8 for each category to product the Potential Non-Ag. Clients. There is a total of 59,031,378 owner-occupied households in the US, and a total of 59,031,378 Potential Non-Ag. Clients. The procedure was confirmed with John Long, Maryland. The final numbers were checked against Maryland's. Two consistent differences were found between the two sets of numbers. First, the White male and female households, and the Total Households per county are greater than John's Maryland calculations by the number of Other Non-Hispanic households. We added them to the Whites while John deleted them. Also the number of Asian male and female households do not balance with the older Maryland figures, although the total Asian households to balance. John may have used the total male/female ratio instead of the Asian male/female ratio to allocate. The Hispanic, Black, and Native American numbers match precisely. The query FinalCountyData has the twenty demographic counts for each county, ten for Ag. Clients, and 10 for non-Ag. Clients. This potential clients work was later duplicated step by step with zip code data. Merging the Ag. Census with the CensusThe next step merged the number of Ag. Clients with the Non-Ag. Census counts in the query Total Clients by County_ Ag Clients_Farms_NonAgClients. This query was exported to the spreadsheet CoClient.xls for distribution. Text files were also developed for non-Excel users. There were a few counties appearing in one dataset, but not the other, particularly Alaska. Both sets of counties were used so the local office can do the sorting. The Zip code data was also merged. There are 7,000 zip codes in the 1992 Ag. Census but not in the 1990 Population Census. There are 1,200 zip codes in the Population Census that are not in the Ag. Census (expected since there are inter-city areas without farm owners). The zip codes in the Population Census are based on a fixed GIS zip code distribution developed for the 1990 Population Census.The Zip code in the Ag. Census is whatever zip code the 'operator/owner' used as his HOME address. The US Postal Service continually revises zip codes for their ongoing distribution systems. The zip codes in most GIS systems do not completely correspond to any other level, however they come close. The first two digits relate to a US Postal Service distribution center, and the last three to specific postal routes. Most zip codes are in close geographically to their numeric neighbors. Local field offices and state GIS coordinators can use this data. Estimated Acreage by County, Race and SexThe published Ag. Census acreage for each county by sex, by race, and by Hispanic origin is published in the following spreadsheet:This Document Requires
Microsoft Excel Due to consistent requests, a standardized estimation of acres for FOCS/NIMS for each county by race and sex was created in the following document: This Document Requires
Microsoft Excel There are two million acres (out of 945 million acres total farmland) are not allocated by race. Tables 37 and 38 do not disclose acreage information if there are few of a minority categories in a county. Local users will see these cases as they input the data, and are encourage to use their local knowledge to estimate acreage for these small minority categories. 1990, 1997, and Estimated 2002 Population1990, 1997, and estimated 2002 population by total, sex, race, and Hispanic was copied from the US Census Bureau for use in estimating changed since the 1990 census. Local uses are encouraged to modify their information in fast growing or declining areas.File Development and DistributionSee our Data Downloads Page. |
|
|
acreage.xls
Back to Top
Back to Top