Kentucky Equine Methodology Report - 2012

METHODOLOGY
Purpose
The purpose of the Kentucky Equine Survey (KyES) was to describe the equine industry of
the state in terms of demographics of both equids and equine operations (Phase 1) and
economic contribution (Phase 2). No comparable study had been performed since 1977,
thus current information was needed for immediate use; the study will also be used as a
benchmark to measure future change in the industry. The methodology for both phases of
the study is described below.
Phase 1: NASS Inventory Study
The KyES was designed and conducted through a collaborative effort between the
University of Kentucky, the United States Department of Agriculture National Agricultural
Statistics Service (NASS) and the Kentucky Horse Council.
When the study began, no current or comprehensive list of equine operations that included
operations not fitting the NASS definition of a farm existed for Kentucky (according to
NASS-USDA, a farm is any establishment that has at least $1,000 in cash receipts annually).
More specifically, there was little information on equine residing on private residences that
are not used for business purposes. In addition, the long period of time since the previous
equine survey necessitated a communications campaign to inform members of the equine
industry about the survey, why it was beneficial and ensuring individuals that responses
were confidential. List building for the survey sample involved acquiring names and
addresses of members of cooperating equine organizations and a general solicitation for
individuals to submit contact information through web pages hosted by UK and the
Kentucky Horse Council. Additional names and addresses were collected at a series of 36
public engagement meetings. These meetings were programmed by UK faculty and
extension workers in counties around the state. In addition to a presentation about the
survey, an educational program about a horse-related topic was usually provided to
motivate audience attendance. During meetings, attendees were encouraged to provide
contact information to the survey personnel.
The list building efforts resulted in the collection of 13,059 names and addresses. As they
were received, names were checked for duplication against those already on the NASS list
(predominantly operations fitting the definition of a farm) and duplicates were removed. A
portion of operations were contacted by telephone by NASS to obtain preliminary
information regarding the numbers of horses at those operations so that the survey sample
could later be stratified by size (where size of the operation is measured by the
approximate number of horses). The final list was comprised of operations and individuals
ranging including private owners of one to two horses at their residences, boarding
facilities, large commercial breeding operations, and race tracks. From the entire list, a
random sample, stratified by geographic location and size, was drawn and surveys were
sent to 15,000 equine operations. If surveys were not returned, telephone enumerators

contacted the operations to obtain the information. In addition, field enumerators visited
some of the largest farms included in the study to assist with data collection. To capture
information on equine operations not on the list, the equine survey was included in the
Agricultural Coverage Evaluation Survey, which was combined with the June Area Survey
sample in constructing the area component of the sample. Two hundred seventy-nine
segments of land were canvassed by field enumerators who collected data on all
agricultural activities in those areas.
Of the 15,000 surveys distributed, 10,753 (72%) produced responses. Of those, 1,042
refused to participate; the remaining 9,711 records were used for analysis. Surveys from
operations with at least one equid were reviewed, edited and entered into a database by
NASS personnel. When a survey was partially completed or the non-respondent was an
extremely large operation, imputation was utilized to account for non-response. Otherwise,
non-response was accounted for through an adjustment to the original sampling weights.
List sample records were expanded by strata and summarized, then records from the 279
area segments that were not on the list (NOL) were expanded and added to the results of
the list to produce state level multiple frame indicators. To produce more robust county
level indicators, a final reweighting was then done, by which weights on NOL records were
set to zero while weights on list records were adjusted, such that the expanded state list
indication equaled the expanded state multiple frame indication. The list sample records
were expanded by this final weight to produce county level indications. The estimation
process produced an estimate of total equine in the state with a relative error of
approximately 1.2% of the estimate.
Phase 2: Economic Impact Analysis and other Economic Studies
A variety of economic measures of Kentucky’s equine industry was estimated through the
use of data from the NASS inventory study using IMPLAN as well as data collected in a set
of supplementary surveys. The methodology for each of these is described in separate
sections below: (1) IMPLAN analysis; (2) Event Attendance Surveys; (3) Racetrack
Management Survey; and (4) Non-market Valuation Survey.
IMPLAN Analysis
To estimate the economic impact of the equine industry on Kentucky, income and
expenditure data from the NASS inventory study, as well as supplementary data from
studies described below, were utilized in an input-output (IO) model with 2011 IMPLAN
data.
Economic impact is usually measured in three ways:
 The output effect: measures the increase in sales due to the presence of an
industry
 The employment effect: measures the number of jobs generated as a result of the
presence of the equine industry



The value added effect: a measure of new income paid to workers, profits earned
by businesses, or dividends paid to shareholders; in other words, it measures the
amount earned by an individual or business after accounting for explicit and implicit
costs

In each of these three measures, the full economic impact of the equine industry includes
the “multiplier effect,” which summarizes the total impact that can be expected from a
change in a given economic activity. For example, a new breeding facility represents an
economic change which can spur ripple effects or spinoff activities, such as veterinary
services and transportation activities. Multipliers measure the economic impact of these
new products or services, including the resulting spinoff activities.
While there are several types of multipliers, the Type II multiplier is most widely used in IO
analysis. A Type II multiplier includes the effect of direct or initial spending, indirect
spending or businesses buying and selling to each other, and household spending based on
the income earned from the direct and indirect effects. Essentially, these latter induced
effects represent employee spending on goods and services.
All industries included in the analysis are identified in Table 1. Note that banking, legal and
accounting services sectors are not included as data are not available. Furthermore, no
tourism impacts related to the equine sector are included.
Table 1. Equine sectors included in the economic impact analysis and revenue contribution
IMPLAN
Sector
2

Industry Description

Equine Contribution

Grain farming

$

78,412,500

14

Animal production, except cattle, poultry and eggs

$

521,000,000

19

Support activities for agriculture and forestry

$

491,000,000

31

Electric power generation, transmission and distribution

$

16,340,000

33

Water, sewage and other treatment and delivery systems

$

16,340,000

39

Maintenance and repair construction of nonresidential structures

$

149,770,000

42

Other animal food manufacturing

$

26,137,500

203

Farm machinery and equipment manufacturing

$

35,500,000

319

Wholesale trade businesses

$

53,350,000

328

Retail Stores - sporting goods, hobby

$

21,870,000

332

Transport by air

$

2,380,000

335

Transport by truck

$

21,420,000

357

Insurance carriers

$

31,570,000

360

Real estate establishments

$

12,320,000

377

Advertising and related services

$

5,135,000

379

Veterinary services

$

58,220,000

403

Spectator sports companies

$

212,938,470

425

Civic, social, professional and similar organizations

$

28,342,500

Separate multipliers were estimated for each of the equine sectors and economic impact
types and are identified in Table 2. The output multiplier, which is used to assess the
interdependence of sectors in the local economy, estimates the total change in local sales

resulting from a $1 increase in sales outside of the study area. Multiplying the increase in
sales of the industry by the output multiplier provides an estimate of the total increase in
sales for the study area, including the initial $1. The employment multiplier measures the
total change in employment resulting from an initial change in employment in the equine
industry. Finally, the value added multiplier provides an estimate of the additional value
added to the product as a result of the equine industry. Value added includes employee
compensation, indirect business taxes, and proprietary and other property income.
Table 2. Estimated multipliers for equine sectors
IMPLAN
Sector

Grain farming

1.591

Value
Added
Multiplier
1.534

Animal production, except cattle and poultry and
eggs
Support activities for agriculture and forestry

1.559

1.841

2.502

1.182

1.758

1.551

1.247

1.164

1.486

1.401

1.962

3.446

1.679

1.674

1.890

2.205

1.698

1.662

1.415

1.473

42

Electric power generation, transmission and
distribution
Water, sewage and other treatment and delivery
systems
Maintenance and repair construction of
nonresidential structures
Other animal food manufacturing

1.453

2.871

3.191

5.176

203

Farm machinery and equipment manufacturing

1.433

1.649

1.847

2.796

319

Wholesale trade businesses

1.520

1.402

1.408

1.693

328

Retail Stores - sporting goods, hobby

1.673

1.606

1.433

1.253

332

Transport by air

1.633

1.624

1.539

2.038

335

Transport by truck

1.748

1.789

1.561

1.677

357

Insurance carriers

1.466

1.407

1.555

2.077

360

Real estate establishments

1.358

1.287

2.116

1.349

377

Advertising and related services

1.521

1.439

1.433

1.532

379

Veterinary services

1.619

1.557

1.356

1.299

403

Spectator sports companies

1.977

2.212

1.731

1.449

425

Civic, social, professional and similar organizations

1.813

1.716

1.345

1.288

2
14
19
31
33
39

Industry Description

Output
Multiplier

Labor
Income
Multiplier
1.792

Employment
Multiplier
1.214

Event Attendance Surveys
Equine-related events generate a number of tourism impacts. In addition to competitors,
family and friends as well as pure spectators attend these events. They spend money on
the event grounds, in the local community, and in the state. While total attendance at
equine events is usually not tracked, we attempted to estimate per-person spending at
events. A team of researchers from the Kentucky Equine Survey attended a number of
events and race meets and intercepted attendees throughout the day of the event.
Attendees were asked to complete a questionnaire which included demographic
information as well as questions related to spending in different categories in different
areas of the state.
All attempts were made to broadly sample different types of equine events, including major
horse shows, smaller horse shows, and organized trail rides. In addition, the study team

visited race meets across the state. Details regarding all events and race meets attending
are identified in Tables 3 below.
Table 3. Horse Events and Race Meets Selected for Event Attendance Surveys
Event

Breed/Discipline

County

Date

# of surveys
collected

Western KY Open Horse Show

All Breeds

Henderson

6/23/2012

48

Region 14 Arabian Championship

Arabian

Fayette

6/30/2012

76

State 4H Horse Show

All Breeds

Jefferson

7/5/2012

76

Churchill Downs

Thoroughbred Racing

Jefferson

7/16/2012

237

Clayton Woosley HOF Reining

Quarter Horse

Fayette

7/28/2012

77

Shelbyville Horse Show

All Breeds

Shelby

8/3/2012

79

Mid-Summer Horse Show

Tennesee Walker

Lincoln

8/11/2012

61

World's Championship Horse Show

Saddlebred

Jefferson

8/25/2012

80

KQHA Fall Futurity

Quarter Horse

Casey

9/1/2012

75

Kentucky Downs

Thoroughbred Racing

Simpson

9/8/2012

213

NKHN Trail Ride

All Breeds

Campbell

9/15/2012

46

Knott County Trail Ride

All Breeds

Knott

10/6/2012

75

The Red Mile

Standardbred Racing

Fayette

10/7/2012

250

Keeneland

Thoroughbred Racing

Fayette

10/27/2012

250

National Horse Show

Hunter/Jumper/Saddlebred

Fayette

11/3/2012

79

Using the data from these surveys, researchers were able to estimate per-person spending
at each event as well as total economic impact from the event (categorized by size).
Racetrack Management Survey
A survey was sent to the general manager or owner of six racetracks in Kentucky: Churchill
Downs, Ellis Park, Keeneland, Kentucky Downs, Red Mile, and Turfway Park. The survey
requested information from the calendar year 2011 in the following categories: revenues,
operating expenses, assets, capital investments, investments in human capital and
technology, and use of wagering technology. Despite repeated attempts to get the surveys
completed, only one was returned. Hence, all needed racetrack data had to be obtained
from the Kentucky Horse Racing Commission’s 2010-2011 biennial report
(http://khrc.ky.gov/reports/Biennial%20Report%202010-2011.pdf).
Non-market Valuation Survey
The non-market valuation study was conducted to obtain an estimate of the value of the
externalities generated by the presence of the equine industry in the state of Kentucky,
which may include recreational, environmental, and aesthetic benefits. The writing and
administration of the survey was accomplished in four stages. First, a preliminary draft of
the survey was created to closely replicate the survey used in Richard Ready’s 1990 study
“The Value to Kentuckians of the Kentucky Equine Industry: A Contingent Valuation Study.”
Second, a focus group was conducted to examine the effectiveness, clarity, and navigability

of the survey instrument. Third, the final draft of the survey was prepared; 6,176 survey
instruments were compiled and mailed to eight counties (Bourbon, Clark, Fayette,
Harrison, Jessamine, Madison, Scott, Woodford) in the Bluegrass Region of Kentucky, and
an additional 2,000 surveys were mailed to randomly selected Kentucky residents outside
of the Bluegrass Region.
A preliminary draft of the survey was created using the survey methodology section and
body of the paper “The Value to Kentuckians of the Kentucky Equine Industry: A
Contingent Valuation Study.” Unfortunately, a copy of the contingent valuation survey used
in the 1990 study was not included in the appendix of the report as stated in the text. To
allow for comparability of the two studies, the explanation of the 1990 contingent valuation
survey (found in the body of the paper) was used to replicate the original survey as closely
as possible. It was tested for effectiveness, clarity, and navigability during a focus group
held at Southside Christian Church of Lexington, KY on August 29, 2012.
The questionnaire was distributed to 8,176 households throughout the state of Kentucky in
two separate mailings. Table 4 presents the number of surveys distributed by county.
Fayette county respondents were randomly selected from a database obtained from the
Fayette County Property Valuation Administrator. The addresses for all respondents
excluding Fayette County residents were obtained from the company USA Data. Reminder
postcards were mailed to increase response rate. The overall response rate, after
accounting for bad addresses, was 22.28%. As expected, response in the Bluegrass Counties
was significantly higher than non-Bluegrass Counties (25.19% vs. 10.10%).
Table 4: Survey Distribution by County
County

Surveys
Distributed

County

Surveys
Distributed

County

Surveys
Distributed

Adair (NB)

12

Grant (NB)

12

Mason (NB)

9

Allen (NB)

10

Graves (NB)

21

Meade (NB)

15

Anderson (NB)

12

Grayson (NB)

15

Menifee (NB)

3

Ballard (NB)

5

Green (NB)

7

Mercer (NB)

12

Barren (NB)

24

Greenup (NB)

22

Metcalfe (NB)

6

Bath (NB)

6

Hancock (NB)

5

Monroe (NB)

6

Bell (NB)

14

Hardin (NB)

54

Montgomery (NB)

14

Boone (NB)

62

Harlan (NB)

15

Morgan (NB)

6

Bourbon (B)

216

Harrison (B)

203

Muhlenberg (NB)

16

Boyd (NB)

27

Hart (NB)

10

Nelson (NB)

23

Boyle (NB)

16

Henderson (NB)

24

Nicholas (NB)

4

Bracken (NB)

5

Henry (NB)

8

Ohio (NB)

12

Breathitt (NB)

9

Hickman (NB)

3

Oldham (NB)

29

Breckinridge (NB)

10

Hopkins (NB)

25

Owen (NB)

5

Bullitt (NB)

41

Jackson (NB)

7

Owsley (NB)

0

Butler (NB)

7

Jefferson (NB)

408

Pendleton (NB)

7

Caldwell (NB)

7

Jessamine (B)

524

Perry (NB)

15

Calloway (NB)

19

Johnson (NB)

13

Pike (NB)

36

Campbell (NB)

48

Kenton (NB)

85

Powell (NB)

7

Carlisle (NB)

3

Knott (NB)

8

Pulaski (NB)

35

Carroll (NB)

6

Knox (NB)

16

Robertson (NB)

1

Carter (NB)

15

Larue (NB)

8

Rockcastle (NB)

8

Casey (NB)

9

Laurel (NB)

32

Rowan (NB)

12

Christian (NB)

32

Lawrence (NB)

9

Russell (NB)

10

Clark (B)

384

Lee (NB)

4

Scott (B)

509

Clay (NB)

10

Leslie (NB)

6

Shelby (NB)

21

Clinton (NB)

5

Letcher (NB)

13

Simpson (NB)

10

Crittenden (NB)

5

Lewis (NB)

6

Spencer (NB)

9

Cumberland (NB)

4

Lincoln (NB)

13

Taylor (NB)

14

Daviess (NB)

52

Livingston (NB)

6

Todd (NB)

6

Edmonson (NB)

6

Logan (NB)

15

Trigg (NB)

8

Elliott (NB)

2

Lyon (NB)

4

Trimble (NB)

5

Estill (NB)

8

Madison (B)

895

Union (NB)

8

Fayette (B)

3176

McCracken (NB)

37

Warren (NB)

55

Fleming (NB)

8

McCreary (NB)

8

Washington (NB)

6

Floyd (NB)

21

McLean (NB)

5

Wayne (NB)

10

Franklin (NB)

27

Magoffin (NB)

6

Webster (NB)

7

Fulton (NB)

4

Marion (NB)

10

Whitley (NB)

19

Gallatin (NB)

4

Marshall (NB)

18

Wolfe (NB)

4

Garrard (NB)

9

Martin (NB)

5

Woodford (B)

269