How are ARMS Data Collected? an Overview

1
How are ARMS Data Collected?an Overview

• Rich Allen
• Deputy Administrator
• National Agricultural Statistics Service

2
Goals

• Clarify survey content
• Describe data interrelationships
• Outline sampling approach
• Mention publicity approaches
• Highlight data collection concerns
• Touch on analysis/summary

3
FarmProductionExpendituresSurvey1972 - 1984
4
Cost ofProductionSurveys1973 - 1984
5
FarmCosts andReturnsSurvey1985 - 1995
6
CroppingPracticesSurvey1990 - 1995
7
AgriculturalResourceManagementSurvey1996
2???
8
Multiple phases are used for survey efficiency,
timeliness, and data linkages
9
ARMS Phase I

• Conducted between May and July to screen farms
  for the later phases
• Ensures operations are still in business
• Defines the sampling frame for later phases
• Shortens initial contact (average 5-10 minutes)
• Collects general farm data such as crops grown,
  livestock inventory, and value of sales

10
ARMS Phase II

• Conducted from September through December when
  information should be fresh
• Collects data on chemical usage, production
  practices, resource use, and variable costs of
  production for specific commodities
• Shortens interview by focusing on specific
  commodities (3-year average 60 Minutes)

11
ARMS Phase III

• Conducted from February through April when
  records should be complete
• Collects information on whole farm finance,
  operator, and household characteristics
• Uses multiple questionnaire versions
• Requires a considerable amount of time (3-year
  average 94 minutes)

12
Cost of Production Matrix 1
Year Commodities
2000 Dairy, Rice, Sugar beets
2001 Corn
2002 Soybeans
2003 Barley, Upland Cotton, Sorghum
2004 Peanuts, Durum Wheat, Other Spring Wheat, Winter Wheat, Hogs
13
Cost of Production Matrix 2
Year Commodities
2005 Corn, Dairy, Oats
2006 Soybeans, Poultry
2007 Upland Cotton, Rice

14
2003 ARMS II Sample Sizes
Commodity Samples
Barley 2,133
Sorghum 1,433
Cotton 2,559
Positive responses from Phase II move to Phase
III.
15
2003 ARMS III Typical Versions
Version Samples
Cost Returns Report 12,375
Barley PPCR 2,133
Sorghum PPCR 1,433
Cotton PPCR 2,559
Subtotal 18,500
(Core 15 states 16,850)
16
Chem Use Matrix 1
Year Commodities
2000 Corn, Upland Cotton, Soybeans, Durum Wheat, Other Spring Wheat, Winter Wheat
2001 Upland Cotton, Soybeans, Fall Potatoes
2002 Corn, Durum Wheat Wheat, Other Spring Wheat, Winter Wheat
17
Chem Use Matrix 2
Year Commodities
2003 Corn, Fall Potatoes
2004 Soybeans
2005 Upland Cotton, Fall Potatoes
2006 Durum Wheat, Other Spring Wheat, Winter Wheat
2007 Corn, Fall Potatoes
18
2003 ARMS III Expansion

• New funding provides for increased sample sizes
  for state level data
• A shorter core questionnaire has been designed
• Two survey modes will be used for initial
  contacts
• Personal
• Mail

19
Sample Selection Objectives

• Represent all operations
• Create detailed data sets
• Minimize respondent burden

20
Perry Burt Sampling

• Sample selection weights adjusted for previously
  selected operations
• Approach virtually eliminates chance of selection
  in consecutive years

21
Stratification

• Main variable is economic size class
• Sampling rates adjusted for operations selected
  for commodity specific contacts

22
Data Expansions

• Basic weights based on inverse of selection
  factors
• Weights adjusted within strata for missing data

23
Data Edits and Analysis

• Records edited for internal consistency
• Individual expansions examined for outliers
• National and Regional summaries are reviewed

24
Routine ARMS III Uses

• Income and expense estimates to BEA - - late June
• Farm Production Expenditures Report (NASS) - -
  mid-July
• National Income Estimates (ERS website) - -
  September
• Ag Income and Finance S O (ERS website) - -
  November

25
ARMS Data Collection Concerns

• Survey seems long
• Some questions are not farm related
• Some question seem quite personal
• Respondents often dont understand Whats in it
  for me?
• Some questions ask for things that cant be
  easily answered

26
Past Data Collection Approaches

• Strengthen enumerator training
• Minimize overlap with other surveys
• Send out pre-survey letters
• Set up interview appointments
• Send back personalized summary information

27
Survey Promotion Approaches

• NASS awareness postcards
• Pre-survey theme letters
• Data user testimonies
• Promotional brochures

28
Data Collection Help Needed

• How to stress Whats in it for you?
• How to describe state and local data uses