National Academies Press: OpenBook

Improving Crop Estimates by Integrating Multiple Data Sources (2017)

Chapter: Appendix A: NASS County-Level Survey Programs

« Previous: References
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

Appendix A

NASS County-Level Survey Programs

COUNTY CROP ESTIMATES1

The quarterly crops Acreage, Production, and Stocks (APS) surveys collect data on crop acreage, yield, and production by commodity and quantities of grain and oilseeds stored on farms used to develop state- and national-level estimates. Since 2011, the fully implemented annual County Agricultural Production Survey (CAPS) has collected data on crop acreage, yield, and production by commodity from a supplemental probability sample of farms; these data are designed to be used with the September and December APS list samples to produce county-level estimates.

The annual APS cycle begins with the June Area Survey (JAS), an area sample described in more detail in the next section. List samples are conducted in June, September, December, and March, with the survey content differing each quarter to capture the seasonality of agriculture: the June survey collects data on planted acres for spring-planted crops and acres harvested and to be harvested for spring crops and winter wheat; the September survey collects data on final harvested acres and production (or yield) of small grains; the December survey collects data on seeded acres of winter wheat (new crop) and final harvested acres and production (or yield) for row crops; and the March survey collects data on winter wheat acres to be harvested as grain, in addition to planting intentions for the spring-planted crops. Final state-level estimates of planted and harvested area,

___________________

1 This appendix draws heavily on information provided on the NASS website as well as information provided to the panel at its meetings by NASS staff. This appendix has been fact-checked by NASS.

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

production, and yield for small grains are published in the Small Grains Annual Summary each September; final national- and state-level estimates of acreage, production, and yield for major row crops are published in the Crop Production Annual Summary each January.

NASS’s county crop estimates program is defined jointly by NASS, the U.S. Department of Agriculture’s (USDA) Risk Management Agency (RMA), and USDA’s Farm Service Agency (FSA). Forty-three states are partners in the process and may add commodities to the program to cover special needs of local cooperators provided external funding. Currently, small grain county estimates are published for 37 states, while row crop county-level estimates are published for 43 states. As of 2016, the list of federal program commodities included

  • barley,
  • dry edible beans,
  • corn—for grain and for silage,
  • cotton—upland and pima,
  • flaxseed,
  • hay—alfalfa and other,
  • oats,
  • peanuts,
  • potatoes,
  • rice,
  • sorghum,
  • soybeans,
  • sugarcane for sugar,
  • sugarbeets,
  • sunflower,
  • tobacco, and
  • wheat—Durum, other spring, and winter.

Survey Timeline

The data collection period for each APS survey lasts roughly 15 days, beginning near the first of the month in which the survey is conducted. The first of the month serves as a reference date for all questions regarding stocks; acreage and production data are reported based on the date of the interview. Thus, interviews for the September and December APS surveys capture end-of-season acreage and production data for small grains and row crops, respectively.

Like the APS surveys, CAPS collects data shortly after harvest so that final production and yield are known, yet close enough to harvest so that memory bias is not an issue. Each CAPS survey (small grains and row

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

crops) is administered across two tiers or groups of states, with data for the first tier being collected earlier than data for the second tier. The first-tier states are southern states whose harvest period is earlier than that of northern states. Data collection for Small Grains CAPS is completed by mid-October (subsequent to the publication of state-level estimates in the Small Grains Annual Summary). Row Crops CAPS also is split into two tiers based on the same rationale. Data collection for Row Crops CAPS is completed by mid-January, but only after the release of state-level estimates in the Crop Production Annual Summary. County-level estimates of acreage, yields, and production are released annually. County-level estimates of acreage and production of small grains are released in mid-December. Federal row crop estimates are published in late February through October, depending on the commodity.

Sampling

The target population for the Row Crops APS is all agricultural operators with cropland and/or storage capacity. NASS uses the dual-frame approach noted above, consisting of four APS list frame samples conducted in June, March, September, and December and an area frame sample of operators that are not on the list (NOL) of operators eligible to be selected from the APS sampling population. The target population for CAPS—a list-only survey—is operators with cropland and/or storage capacity in the 43 collaborating states. The NASS list frame includes all known agricultural establishments. The list frame for CAPS consists of those NASS list frame records with positive planted acres or storage capacity of the desired commodities in the previous year.

The Row Crops APS and CAPS list frame samples are selected using a multivariate probability proportional to size (MPPS) sampling scheme in which the measure of “size” is determined by more than one item (see Bailey and Kott, 1997). The MPPS design allows target sample sizes for the commodities of interest to be set at the county level. The probability of selection is determined by taking the maximum (over all targeted commodities) of products involving

  • the targeted sample size for each commodity, and
  • the sampling unit’s size with respect to each targeted commodity—a proportion based on that unit’s planted area relative to the county’s planted area for each commodity.

Calibration of sampling weights to list frame totals is performed.

The sampled records from the list portion of the Row Crops APS are used in the production of state- and national-level estimates, and they

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

are used again in conjunction with CAPS records to produce county-level estimates. Hence each sample record has two different MPPS weights. The first is used in the estimation process for the Row Crops APS surveys at the state and national levels. The second is used to produce estimates for the combined Row Crops APS and CAPS surveys at the county level (both in composite estimators that use the JAS data); effectively, the sampled APS and CAPS records from the list frame are reweighted to represent a single probability sample.

The target coefficient of variation (CV) for acreage is 1–4 percent at the national level and 5–10 percent at the state level. As of 2016, the list-only sample sizes were approximately 59,000 for the September APS survey and 75,000 for the December APS survey. The area NOL sample provided an additional 6,000 records in both September and December to account for undercoverage in the state- and national-level estimates. The sample size was 80,000 for Small Grains CAPS and 171,000 for Row Crops CAPS. When the respective APS and CAPS list samples are pooled, county-level estimates for small grains are based on a sample size of 139,000, and county-level estimates for row crops are based on a sample size of 246,000.

Data Collection

Sampled farms receive a presurvey letter explaining the survey and indicating that respondents will be contacted for survey purposes only. The letter provides the questions that will be asked to allow respondents to prepare in advance, and also provides a pass code they can use to complete the survey online. All modes of data collection are utilized for the September and December Row Crops APS surveys as well as the Small Grains CAPS and Row Crops CAPS. Regional Field Offices (RFOs) have the option of conducting a mail-out/mail-back phase. Most of the data are collected through computer-assisted telephone interviews (CATIs) administered by individual RFOs and Data Collection Centers. Limited personal interviewing is done, generally for large operations or those with special handling arrangements.

Before CAPS data collection is complete, an adaptive data collection strategy is used to prioritize counties for nonresponse follow-up to increase commodity coverage and meet publication standards. There are four priority levels based on individual county crop profiles. The first priority is a county with a crop that is close to meeting publication standards. Second is a county for which the publication standard has not been met for at least one crop but for which estimates were published the previous year. The third priority is remaining counties that do not meet publication standards for at least one crop. Finally, the fourth priority is all other counties that meet publication standards. For this last group, the county is still available

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

for nonresponse follow-up, but the remaining nonresponding farms are at the bottom of the call scheduler’s list.

Nonresponse Adjustment

Reweighting is used to account for nonresponse. Nonresponse weighting groups are based on operation size and type as well as Agricultural Statistics District (ASD), a geographically defined grouping of neighboring counties within a state. The nonresponse weighting groups for the Row Crops APS and CAPS samples are based on the control data items for total cropland; on-farm grain storage capacity; rice acreage (rice-producing states); potato acreage (potato-producing states); and some rare or specialty crops, which depend on the state. These nonresponse weighting groups ensure that operation size and location are taken into account during reweighting. In each nonresponse weighting group, the adjustment is calculated by summing the weights for all sampled records and dividing by the sum of the weights from the completed records. This ratio is applied to the weights of the completed records and assumes that the data for the nonrespondents are similar to the data for the respondents. No item-level imputation is performed.

Estimators

The list-only APS surveys and NOL component obtained from the JAS are used together to estimate national- and state-level acreage, production, and yield. Data from the list-only portions of the APS surveys and CAPS are combined to produce ASD- and county-level estimates. The small grains (wheat, barley, and oats) county-level estimates are direct expansions utilizing the combined Small Grains CAPS and September Row Crops APS list records. Similarly, the county-level estimates for row crops (e.g., corn, soybeans, and cotton) are direct expansions of the combined Row Crops CAPS and December Row Crops APS list records.

Explicit formulae used by NASS may be found in Kott (1989). Two kinds of estimators are used for county-level indications: direct expansions and ratio estimators. Direct expansions, or Horwitz-Thompson estimators, are used to estimate such totals as planted and harvested acres and production. Reweighted direct expansions are calculated by summing the reported commodity values multiplied by the nonresponse-adjusted sample weights in each nonresponse weighting group. All reweighted direct expansions are computed at the state, district, and county levels.

The yield ratio estimator takes the form of a ratio of two reweighted direct expansion estimates—production and harvested acres—computed as described above. Other survey ratio indications are commodity planted

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

acres to total cropland acres and commodity harvested acres to planted acres. In addition, ratio estimators are used for all across-survey ratios (e.g., current-year to previous-year planted acreages). For the survey-to-survey ratios, both the current and previous survey data must be reported (positive or valid zero) to be included in the ratio. If either of these components is not reported, the sampling unit is excluded from the estimate, and the weights of the completed records are adjusted accordingly. All ratio indications are calculated at the state, district, and county levels.

Variances for all total indications are calculated using the delete-a-group jackknife method with 15 replicate groups. Variances for ratio indications are constructed using a second-order Taylor series expansion for the ratio. The variance–covariance structures for the numerator and denominator (totals) are constructed using the delete-a-group jackknife with 15 replicated weights (Kott, 1989).

The Agricultural Statistics Board (ASB) Process for County Estimates

RFO statisticians collect survey data, perform data editing, and conduct further data analysis. Estimates established by RFO staff and appointed headquarters staff are submitted for review to ASB. Official county-level estimates are set by incorporating a variety of input sources, including direct survey estimates of totals and ratios, FSA data on planted acres, a variant of the famed Battese-Harter-Fuller small-area model of planted acreage (Battese et al., 1988), RMA data on failed acres (to improve harvested area totals), and remotely sensed estimates for production and yield for corn and soybeans in selected states. Past official estimates also are reviewed. From these inputs, NASS constructs composite estimates for its planted area, harvested area, and production totals, which serve as the starting point for setting official estimates. Prescribed initial weights are provided by ASB, although these weights may be adjusted as necessary. The composite estimate at the ASD level is ratio benchmarked to the previously published state totals. NASS rounding rules are enforced, generating official ASD estimates. Subsequently, the composite estimates at the county level are ratio adjusted to the rounded ASD estimates. Enforcement of NASS rounding rules at the county level generates the official county-level totals. All official estimates of yield are derived as the ratio of corresponding final estimates of the production and harvested area totals.

ASB reviews estimates established by RFO staff and appointed headquarters staff. The RFO provides justification to headquarters when recommended estimates deviate from survey results. ASB members review all recommended state-, district-, and county-level estimates for accuracy and consistency across state boundaries and verify that proper procedures were

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

followed. Through this review process, ASB has final approval of all official crop estimates nationwide.

NASS verifies that estimates meet publication standards that must be met before any official estimate can be published. First, if the combined sample sizes realized for the CAPS and APS surveys include at least 30 positive reports of crop production, the county-level (or ASD-level) official estimate is suitable for publication. In counties with fewer than 30 but at least 3 positive reports of crop production, a 25 percent coverage condition is verified; these county estimates may still be published provided the corresponding unweighted harvested acreage reports account for at least 25 percent of the official estimate of harvested area set by ASB. Counties with fewer than 3 positive reports of crop production are automatically suppressed. Complementary suppressions may be necessary to avoid disclosing suppressed county estimates. These standards are verified independently for every commodity.

Official estimates are open to revision on a preannounced schedule only if new information becomes available. If changes are made to the state-level official estimates during the normal annual revision period (timing varies by commodity), the county-level data are revised to ensure that county- and ASD-level estimates continue to sum to state-level estimates. These previous-year revisions are released at the same time that the data for the current year are published.

The panel was told that small grain county estimates were published for about 17–53 percent of eligible counties in 2014. The percentage of U.S. total production covered by these published individual county-level estimates of small grains ranged from 52 percent to 82 percent, depending on the commodity. County-level estimates of row crops were published for 29–68 percent of all eligible counties. The percentage of U.S. total production covered by these published individual county-level estimates ranged from 73 percent to 96 percent, depending on the commodity.

CASH RENTS SURVEY PROGRAM

Since 2008, state-level estimates for cash rents have been set based on a combination of the Cash Rents Survey and JAS in the years in which the Cash Rent Survey is conducted. In years when the Cash Rents Survey is not conducted, state-level estimates are set using the JAS. Estimates are published in the first week of August for all states except Alaska and Hawaii. District- and county-level cash rents estimates are published the second week of September, no less frequently than every other year as mandated by the 2014 farm bill.

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

The June Area Survey

The JAS includes questions on cash rental rates in addition to questions on other agricultural topics. The JAS sample is drawn from the NASS area frame, which covers all land in the United States except for Alaska and Hawaii. The JAS sample design is a stratified random design, with strata based on percent cultivation. Within each stratum, primary sampling units (PSUs) of 2 to 8 square miles are selected. The selected PSUs are divided into segments of approximately 1 square mile, and a segment is randomly selected from within the PSU. Once selected, a segment remains in the sample for 5 years. Each year approximately 20 percent of the segments enter the sample, and 20 percent rotate out. Each segment is divided into tracts, each representing a unique operation. The national sample consists of about 11,000 segments comprising some 38,000 tracts.

Data collection for the JAS is conducted by personal interview. Enumerators must account for all operations and land contained in their assigned segments. Enumerators conduct interviews and collect responses from the end of May through mid-June. Survey questionnaires are returned to the RFOs, where they are reviewed visually and entered manually into a database.

For the JAS cash rents, item-level nonresponse is accounted for by imputing data where values are missing. Imputed values are calculated through an automated imputation algorithm that requires a minimum of five complete reports within the imputation group. When a group lacks a sufficient number of responses, groups are collapsed according to a defined hierarchy, preserving as much of the homogeneity as possible, until five complete reports have been identified. The first imputation group is the segment, the second is similar strata within the county, the third is the county, the fourth is the district, and the fifth is the state.2

The JAS uses direct expansion estimates of cash rent items and their respective variances. Because cash rent items pertain to the entire farm, including those portions lying outside the sampled tract, the survey weight for a record is the product of the original segment sampling weight from the area frame and the proportion of the farm residing within the segment boundaries (called the farm-to-tract ratio). Rent per acre is computed as the ratio of the weighted estimate of dollars of rent paid to the weighted estimate of acres rented.

___________________

2 Drawn from USDA, NASS ISSN 2167-129X, “Cash Rents Methodology and Quality Measures.” See https://www.nass.usda.gov/Publications/Methodology_and_Data_Quality/Cash_Rents/08_2016/rentqm16.pdf [October 2017].

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

The Cash Rents Survey

The target population for the Cash Rents Survey is all farms and ranches with $1,000 or more in agricultural sales (or potential sales) that operate land rented from others on a cash basis in counties with at least 20,000 acres in cropland or pastureland (excluding Alaska) based on the most recent Census of Agriculture. The Cash Rents Survey collects data on acres rented and cash rental rates or total dollars paid in rent for three land use categories: irrigated cropland, nonirrigated cropland, and permanent pastureland. The data collection period for the survey is from late February through the end of June.3

The Cash Rents Survey sample is selected from the NASS list frame. The sample design for the survey is a stratified systematic sample at the county level, stratified by irrigated acres, nonirrigated acres, and pastureland. Within strata, farms are sorted by rent expenses. The certainty strata are those with a sampling interval of 1 in Table A-1, usually the strata that include farms making up more than 10 percent of the county total in certain categories. Most of the noncertainty strata use a 1 in 2 sampling rate. The noncertainty stratum for rent paid for all land and buildings uses a 1 in 3 sampling rate. The category of land rented or leased from others uses a sampling interval of 1 in 40. In 2014, the population was 745,489, farms, and the sample size was 241,176.

The Cash Rents Survey sample is drawn in November, and the questionnaire is finalized in January. The survey data are collected early in the year, when farmers know their rental rates. The initial mailing starts in February; calling begins in March; and nonresponse mailing begins in April. Data collection by personal visit,4 to either large operations or those with special handling arrangements, concludes in July. There is also an electronic reporting option. State-level estimates are published in August and county-level estimates in September. Data collection is coordinated with other surveys during this time period to minimize respondent burden. Data are collected on acres (by irrigated cropland, nonirrigated cropland, and pastureland) by acres rented to and acres rented from. Data on rent in dollars are collected for the same breakdown. Data on the total cash rents paid for all land and buildings also are collected. Family rents and cash rents are included in the survey, while share, flex, hybrid, and bonus rents are excluded.

___________________

3 To help with survey coordination in the month of March, the Cash Rents Survey questions are included on the March Agricultural Survey (see http://www.nass.usda.gov/Surveys/Guide_to_NASS_Surveys/Crops_Stocks [October 2015]) questionnaire. And during June, cash rents data are collected on the JAS in all states except Alaska and Hawaii.

4 Data are collected either on a paper form or by a computer-assisted personal interviewing instrument.

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

TABLE A-1 General Stratified Sample Design for Each County

Stratum Description Sampling Interval
98 Any farms with irrigated cropland acres >10% of total 1
96 Any farms with pastureland acres >10% of total 1
94 Any farms with nonirrigated cropland acres >10% of total 1
90 Irrigated cropland acres 1
80 Pastureland acres 2
70 Nonirrigated cropland acres 2
60 Rent paid for all land and buildings 3
50 Unknown cash rent expenses 1
40 Land rented or leased from others 40

SOURCE: Presentation by Jeff Bailey, National Agricultural Statistics Service, to the Panel on Methods for Integrating Multiple Data Sources to Improve Crop Estimates, November 13, 2015, Washington, DC.

The overall response rate to the Cash Rents Survey was about 76 percent in 2014. Once a county reaches the threshold required by the publication standard, attention to improving response is turned to other counties near the threshold, although responses are still accepted. The goal is to publish estimates for as many counties as possible. With respect to the mode by which completed surveys are received, the breakdown is 1 percent by fax; 2.7 percent by Internet; 39.4 percent by mail; 27.6 percent by telephone; and the remainder, about 5.3 percent, by field enumeration. NASS is considering approaches to encouraging response and cognitively testing the form. It hopes to have a more attractive Internet reporting option in the future.

All nonresponse is accounted for by reweighting. Weighting groups are based on ASD and design strata. Variances of totals are calculated using delete-a-group jackknife. Variances of ratios are calculated based on a second-order Taylor series expansion for ratios (see Kott [1989] for details). No item-level imputation is performed.

Cecere and colleagues (2012) and Berg and colleagues (2014) describe an evolution of the NASS cash rents model. The Berg, Cecere, and Ghosh (2014) model has been used to provide indications to ASB since 2013. The model is univariate, summing results from two separate univariate Fay-Herriott models that estimate the average rental rate (dollars/acre) in 2 years and the difference between rental rates (dollars/acre) in the same 2 years. Both models also include as covariates the National Commodity Crop Productivity Index from the Natural Resources Conservation Service (NRCS), a yield index from NASS historical estimates, and the total value

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

of production from the Census of Agriculture. A two-stage benchmarking process applied to the modeled county-level estimates generates ASD-level estimates and satisfies relationships with the published state-level cash rental rates.

As is common with most NASS surveys, RFOs collect the data, resolve record-level errors identified during the editing process, and conduct further analysis within and across records before computing survey estimates. Final estimates are set through an ASB process that involves appointed NASS RFO and Headquarters staff. For cash rents, the primary basis for official estimates is the direct survey estimate. The Cecere, Berg, and Ghosh modeled estimates provide an additional source of information for ASB’s consideration, as do previously reported survey and frame data.

The NASS publication standard for the Cash Rents Survey is that there must be at least 30 positive reports of rent in the county, or if there are fewer responses, the sum of their corresponding unweighted acreage must account for at least 25 percent of rented acreage estimated. Estimates based on samples of fewer than 3 reports are automatically suppressed. While NASS computes estimates for total acres rented and total rent expense in a county, it does not publish these separate series; it publishes only the dollar per acre rental rate. The publication standard is verified independently for each of the three land use categories: irrigated cropland, nonirrigated cropland, and permanent pastureland.

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×

This page intentionally left blank.

Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 111
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 112
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 113
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 114
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 115
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 116
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 117
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 118
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 119
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 120
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 121
Suggested Citation:"Appendix A: NASS County-Level Survey Programs." National Academies of Sciences, Engineering, and Medicine. 2017. Improving Crop Estimates by Integrating Multiple Data Sources. Washington, DC: The National Academies Press. doi: 10.17226/24892.
×
Page 122
Next: Appendix B: Routine External Evaluation Protocol »
Improving Crop Estimates by Integrating Multiple Data Sources Get This Book
×
Buy Paperback | $65.00 Buy Ebook | $54.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The National Agricultural Statistics Service (NASS) is the primary statistical data collection agency within the U.S. Department of Agriculture (USDA). NASS conducts hundreds of surveys each year and prepares reports covering virtually every aspect of U.S. agriculture. Among the small-area estimates produced by NASS are county-level estimates for crops (planted acres, harvested acres, production, and yield by commodity) and for cash rental rates for irrigated cropland, nonirrigated cropland, and permanent pastureland. Key users of these county-level estimates include USDA’s Farm Services Agency (FSA) and Risk Management Agency (RMA), which use the estimates as part of their processes for distributing farm subsidies and providing farm insurance, respectively.

Improving Crop Estimates by Integrating Multiple Data Sources assesses county-level crop and cash rents estimates, and offers recommendations on methods for integrating data sources to provide more precise county-level estimates of acreage and yield for major crops and of cash rents by land use. This report considers technical issues involved in using the available data sources, such as methods for integrating the data, the assumptions underpinning the use of each source, the robustness of the resulting estimates, and the properties of desirable estimates of uncertainty.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!