removed from the list. 
The official CML for the 2012 Census of Agriculture 
was established on September 1, 2012. The list 
contained 3,009,641 records. There were 2,387,326 
records that were thought to meet the NASS farm 
definition and 622,315 potential farm records, which 
included NACS nonrespondents, other records added 
to the CML by the NASS field offices after the 
record linkage process, and late adds to the CML 
that were not included in any previous NACS or 
State screening survey. 
Not on the Mail List (NML) 
Extensive efforts are directed toward developing a 
CML that includes all farms in the U.S. However, 
some farms are not on the list, and some agricultural 
operations on the list are not farms. NASS uses its 
June Agricultural Survey (JAS) to quantify the 
number and types of farms not on the CML. The 
tracts in the JAS that are not on the CML are said to 
be in the Not on the Mail List (NML) domain. If a 
tract in the NML domain is determined to be a farm 
during the census, it is an NML farm. The NML 
farms are used to estimate the undercoverage 
associated with the census. 
The NASS area frame, which is used for the JAS, 
covers all land in the U.S. and includes all farms. 
The land in the U.S. is stratified by characteristics of 
the land. A probability sample of segments is drawn 
within each stratum for the JAS. Segments of 
approximately equal size are delineated within each 
stratum and designated on aerial photographs. The 
JAS sample of segments is allocated to strata to 
provide accurate measures of acres planted to widely 
grown crops, farm numbers, and inventories of 
cattle. Sampled segments in the JAS are personally 
enumerated. Each operation identified within a 
segment boundary is known as a tract. 
The 2012 JAS sample was increased to improve the 
farm counts for operations that produced specialty 
commodities or had socially disadvantaged or 
minority operators. The total sample consisted of 
14,376 segments of which 3,291 were additional 
segments added to facilitate the use of the JAS as an 
Agricultural Coverage Evaluation Survey (ACES). 
The additional segments were added based upon 
A- 2 APPENDIX A 
multivariate sample allocations to target specific 
items at the U.S. level. The 2012 JAS consisted of 
sample segments from all States, with the exception 
of Alaska where NASS does not maintain an area 
frame. 
During the JAS prescreening operation, each tract is 
identified as either agricultural or non-agricultural. 
Each JAS agricultural tract is identified as a farm or 
non-farm in June based on the farm definition. Non- 
agricultural tracts are further classified into 
categories; with farm potential, with unknown farm 
potential, or with no farm potential. The names and 
addresses collected in the 2012 JAS were matched to 
the CML. Those from the JAS 2012 survey that did 
not match were determined to be in the NML domain 
and sent a yellow census report form so that they 
could be differentiated from the green report form 
sent to those addressees on the CML. Instructions 
on the census report form directed any respondent 
who received duplicate forms to complete the CML 
form and to mail all duplicate forms back together. 
Those who returned a CML and an NML form had 
been misclassified as NML and were removed from 
the NML domain. 
The initial NML mailout consisted of 36,021 
records. An additional 403 June area tracts linked to 
Census records that were Undeliverable as 
Addressed (UAA) were later added to the NML 
domain. A total of 36,424 NML records were 
summarized of which 5,565 records were truly NML 
and in-scope. 
The farm/nonfarm status of each NML domain 
operation was determined based on the reported data 
in the census form. An operation in the NML domain 
that was determined to be a farm is referred to as an 
NML farm. Characteristics of NML farms and their 
operators provided a measure of the undercoverage 
of farms on the CML. The percentage of farms not 
represented on the CML varied considerably by 
State. In general, NML farms tended to be small in 
acreage, production, and sales of agricultural 
products. Farm operations were missing from the 
CML for various reasons, including the possibility 
that the operation started after development of the 
CML, the operation was so small that it did not 
appear in any agriculture-related source list, or the 
operation was misclassified as a nonfarm prior to 
census mailout. The CML was used with the NML in 
2012 Census of Agriculture 
USDA, National Agricultural Statistics Service 
