USDA: Direct & Intermediated Marketing of Local Foods US

8/3/2019 USDA: Direct & Intermediated Marketing of Local Foods US

1/38

United StatesDepartment ofAgriculture

EconomicResearch

ServiceEconomicResearchReportNumber 128

November 2011

Direct and IntermediatedMarketing of Local Foodsin the United States

Sarah A. Low

Stephen Vogel


2/38

Visit Our Website To Learn More

About Local Foods Systems!

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its

programs and activities on the basis of race, color, national origin, age,

disability, and, where applicable, sex, marital status, familial status, parental

status, religion, sexual orientation, genetic information, political beliefs, reprisal,or because all or a part of an individuals income is derived from any public

assistance program. (Not all prohibited bases apply to all programs.) Persons

with disabilities who require alternative means for communication of programinformation (Braille, large print, audiotape, etc.) should contact USDAs

TARGET Center at (202) 720-2600 (voice and TDD).

To file a complaint of discrimination write to USDA, Director, Office of Civil

Rights, 1400 Independence Avenue, S.W., Washington, D.C. 20250-9410 or

call (800) 795-3272 (voice) or (202) 720-6382 (TDD). USDA is an equalopportunity provider and employer.

www.ers.usda.gov

Recommended citation ormat or this publication:

Low, Sarah A., and Stephen Vogel.Direct and Intermediated Marketing o

Local Foods in the United States, ERR-128, U.S. Department o Agriculture,

Economic Research Service, November 2011.

Cover photos: Local dairy sign, ERS, USDA. All others, Thinkstock.


3/38

United States

Department

of Agriculture

www.ers.usda.gov

A Report from the Economic Research Service

Economic

Research

Report

Number 128

November 2011

Direct and Intermediated

Marketing of Local Foods

in the United States

Abstract

This study uses nationally representative data on the marketing o local oods to assess

the relative scale o local ood marketing channels. This research documents that salesthrough intermediated marketing channels, such as armers sales to local grocers andrestaurants, account or a large portion o all local ood sales. Small and medium-sizedarms dominate local oods sales marketed exclusively through direct-to-consumer chan-nels (oods sold at roadside stands or armers markets, or example) while large armsdominate local ood sales marketed exclusively through intermediated channels. Farmersmarketing ood locally are most prominent in the Northeast and the West Coast regionsand areas close to densely populated urban markets. Climate and topography avoring theproduction o ruits and vegetables, proximity to and neighboring arm participation inarmers markets, and good transportation and inormation access are ound to be associ-ated with higher levels o direct-to-consumer sales.

Keywords: Local oods, direct marketing channels, direct sales, intermediated sales

Acknowledgments: The authors are grateul to Economic Research Service reviewersRobert Hoppe and Michael Hand; Larry Lev rom Oregon State University; Debra Tropprom USDAs Agricultural Marketing Service; and an anonymous reviewer or insightulcomments and suggestions.

Sarah A. Low, [email protected]

Stephen Vogel, [email protected]


4/38

ii

Direct and Intermediated Marketing o Local Foods in the United States / ERR-128

Economic Research Service / USDA

Contents

Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Farms Marketing Local Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Historical Trends in Direct Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Which Marketing Channels Do Local Food Sales Farms Use? . . . . . . . . . 4

What Commodities Are Being Produced or Local Food Sales?. . . . . . . . 7

Comparing Farms That Market Local Foods With FarmsThat Do Not Market Local Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Location of Local Food Sales Farms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Local Food Sales in U.S. Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0

Local Food Sales Highest in Urban Areas . . . . . . . . . . . . . . . . . . . . . . . . 12

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Appendix 1: Developing Data on Marketing of Local

Agricultural Products Using the 2008 ARMS . . . . . . . . . . . . . . . . . . . . . 18

Appendix 2. Modeling Direct-to-Consumer Sales. . . . . . . . . . . . . . . . . . 23

Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

The Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Hypothesis and Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Empirical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27


5/38

iii

Direct and Intermediated Marketing o Local Foods in the United States / ERR-128Economic Research Service / USDA

Summary

What Is the Issue?

Despite increased production and consumer interest, locally grown oodaccounts or a small segment o U.S. agriculture. For local oods productionto continue to grow, marketing channels and supply chain inrastructure must

deepen. Inormation on U.S. local ood producers and their marketing chan-nels, however, is incomplete. New inormation on armers that market oodslocally and the marketing channels they use presented in this report couldaid private- and public-sector eorts to support this sector o the agriculturaleconomy. This report uses the 2008 Agricultural Resource ManagementSurvey (ARMS) to explore armers use o both direct-to-consumer andinter-mediated marketing channels in selling locally produced oods toconsumers.

What Did the Study Find?

Marketing o local oods, via both direct-to-consumer and intermediatedchannels, grossed $4.8 billion in 2008about our times higher than esti-mates based solely on direct-to-consumer sales.

Farms marketing ood commodities exclusively through intermediatedchannels reported $2.7 billion in local ood sales in 2008over threetimes higher than the value o local oods marketed exclusively throughdirect-to-consumer channels, and two times higher than the value o localoods marketed by arms using a combination o direct-to-consumer andintermediated channels.

Small arms (those with less than $50,000 in gross annual sales) accountedor 81 percent o all arms reporting local ood sales in 2008. They aver-aged $7,800 in local ood sales per arm and were more likely to rely

exclusively on direct-to-consumer marketing channels, such as amersmarkets and roadside stands.

Medium-sized arms (those with gross annual sales between $50,000 and$250,000) accounted or 17 percent o all arms reporting local ood salesin 2008. They averaged $70,000 in local ood sales per arm and werelikely to use direct-to-consumer marketing channels alone or a mix odirect-to-consumer and intermediated marketing channels.

Large arms (those with gross annual sales o $250,000 or more)accounted or 5 percent o all arms reporting local ood sales in 2008.They averaged $770,000 in local ood sales per arm and were equally

likely to use direct-to-consumer channels exclusively, intermediated chan-nels exclusively, or a mixture o the two.

Large arms accounted or 92 percent o the value o local ood salesmarketed exclusively through intermediated channels.

For small and medium-sized arms with local ood sales, more operatorsidentied their primary occupation as arming and devoted more timeto their arm operation than operators o similarly sized arms withoutlocal sales.


6/38

iv



Vegetable, ruit, and nut arms dominated local ood sales.

Direct-to-consumer sales o ood commodities were aected by climateand topography that avor ruit and vegetable production, proximity toarmers markets and neighboring local ood arms, and access to trans-portation and inormation networks.

The value o locally sold ood is highest in metropolitan areas and is

geographically concentrated in the Northeast and on the West Coast.

How Was the Study Conducted?

We used the 2008 ARMS data to analyze armers use ospecifc direct-to-consumer marketing channels (i.e., use o roadside stands, armers markets,onarm stores, and community-supported agriculture arrangements) andintermediatedmarketing channels (i.e., armers sales to local retail, restau-rant, and regional distribution outlets), but also arm characteristics and thevalue o sales or armers engaged in local ood sales. Data rom the 2007Census o Agriculture supported the spatial econometric model used to iden-tiy determinants o direct-to-consumer sales.


7/38

1


Introduction

Although locally grown oods sold by armers directly to consumers, restau-rants, or grocers have become more popular, inormation on local U.S. oodproducers and their marketing channels is incomplete. Case studies haveprovided detailed inormation about local ood sales in certain U.S. regions(e.g., King et al., 2010), but the scope o nationally representative data has

been limited to direct-to-consumer sales. This report provides new inor-mation on marketing channels based on the 2008 Agricultural ResourceManagement Survey (ARMS).

Whether purchased at a armers market or at a nearby grocer, local ood isan ambiguous characteristic o consumer purchases. In this study, the deni-tion o local ood is based on the set o marketing channels (as measuredrom the armgate to the consumer) used by armers (Hand and Martinez,2010; Martinez et al., 2010). Thus, direct-to-consumerand intermediated(direct-to-grocer/restaurant) ood sales are considered local oods in thisstudy.1 National data on direct-to-consumer ood sales have been availablesince the 1978 Census o Agriculture, long beore the current surge in local

oods interest. Nationally representative data on intermediated sales by armoperators, however, became available in the 2008 ARMS. In this report, weexplore the use o both direct-to-consumer and intermediated sales by sizeo arms, their commodity specializations, and the characteristics o armoperators participating in local oods marketing channels. Data rom the2007 Census o Agriculture allow us to explore regional patterns in direct-to-consumer sales.

1Martinez and others (2010) explored

the conceptual and policy cross-cur-

rents embedded in dening local oods.Dening local oods through armers

marketing channels avoids problem-

atic denitions based on geographic

distances between producers and

markets and is congruent with relevant

literature.


8/38

2



Farms Marketing Local Foods

Farmers selling ood commodities through armers markets, roadside stands,and other local ood sales outlets may account or a small segment o U.S.agriculture, but this segment has experienced recent growth and increasedpopularity. Historically, nationally representative data on ood sold directlyto consumers by armers has been used to characterize arms participating in

such sales.

Historical Trends in Direct Sales

Data on direct-to-consumer ood sales were rst collected in the 1978 Censuso Agriculture, ater the Farmer-to-Consumer Direct Marketing Act waspassed, and direct-to-consumer sales data have been collected every 5 yearssince (except or 1987).

Over the 1978-2007 period, arms with direct-to-consumer ood sales repre-sented an average 5.5 percent o all arms, and total direct-to-consumer salesaccounted or 0.3 percent o total arm sales. The number o armers engaged

in direct-to-consumer sales peaked in 1982, likely due to the 1976 Farmer-to-Consumer Direct Marketing Act, which provided unding or activities thatostered direct marketing, such as technical assistance via agricultural exten-sion (g. 1). Between 1992 and 2007, the number o armers participating indirect-to-consumer sales increased by 58 percent to 136,000, and the constantdollar value o direct sales increased by 77 percent to $1.2 billion.2

The Census o Agriculture contains a limited amount o data on direct-to-consumer salesthe number o arms engaged in this activity and the valueo direct-to-consumer sales. Researchers have called or the census to collectinormation on dierent local ood marketing channels (Brown, 2002; Lev

2In 1997, USDA adjusted census data

to correct or small arms previously

missed and, consequently, part o the

1997 increase may be due to changes in

data collection and weighting proce-

dures.

1978

Farms selling local foodsdirectly to consumers Direct sales

of local foods

1982 1992 1997 2002 2007

60

80

100

120

140

160

200

400

600

800

1,000

1,200

1,400

Note: Inflation adjusted sales were calculated based on the gross domestic product implicit pricedeflator published by the Bureau of Economic Analysis, U.S. Department of Commerce andcalibrated to 2007=100.Source: 1978, 1982, 1992, 1997, 2002, and 2007 U.S. Censuses of Agriculture.

Census year

Thousand farms Million dollars (constant 2007)

Figure 1

Direct-sales farms and direct sales of local foods, 1978-2007


9/38

3


and Gwin, 2010). The 2008 ARMS collected this inormation, and these datasuggest that prior data collection eorts that ocused on direct-to-consumersales missed a relatively large portion o local ood salessales o ood orhuman consumption to grocers and restaurants (see box, What Can WeLearn From the 2008 ARMS Data on Local Food Sales?).

Using the 2008 ARMS, we measured direct-to-consumersales o local oodsthrough our outlets: armers markets, roadside stands, onarm stores, and

community-supported agriculture arrangements (CSAs).3Intermediatedmarketing channels include sales to regional distributors and grocery stores,restaurants, or other retailers.4 Generally, marketing channels are classiedas intermediated when local ood products pass through one or more inter-mediate steps in the local ood supply chain beore reaching the consumer(King et al., 2010). While the 2008 ARMS collected data on armers use ospecic local ood marketing channels, it did not collect data on the valueo sales linked to a specic channel. Without these data, researchers mustadhere to the strict trichotomy that groups local ood sales by exclusive useo direct-to-consumer outlets, exclusive use o intermediated channels, ormarketing through both channels (g. 2).

The 2008 ARMS estimates shed light on two characteristics o localood supplies.

Gross sales o locally marketed ood (to consumers and local interme-diaries) are our times larger than previous census and ARMS estimatessuggested, representing 1.9 percent o total gross arm sales, primarilybecause intermediated sales were included or the rst time.

Most local oods are marketed through intermediated channels, accountingor 50-66 percent o the value o all local ood sales.

Our ndings validate local oods researchers concerns that direct-to-

consumer sales account or only a relatively small portion o total local oodsales (Lev and Gwin, 2010; Clancy and Ruh, 2010).

3A CSA buying club is a marketing

arrangement in which a group o house-

holds agree to purchase shares o a

armer's expected yield beore planting.

These upront cash payments allow thearmer to buy inputs and share the out-

put and yield risks with CSA members.

These arrangements are reerred to as

subscription agriculture. Some CSAs

tie households to ormal contracts and

others to inormal arrangements and/

or barter.

4Regional local ood distributors make

up a very small portion o intermedi-

ated local ood sales, suggesting that

armers perceive the dierence between

regional local ood distributors and

national distributors. Also, the 2008

ARMS did not explicitly collect data on

emerging institutional outlets, such as

arm-to-school arrangements.

Direct-to-consumeroutlets

Intermediatedmarketing channelsBoth

$4.8 billion in sales

Figure 2Farmers local food marketing, 2008

71,200 farms$877 million in sales

Farmers markets

Roadside stands

Farm stores

CSA arrangements

13,400 farms$2.7 billion in sales

Grocers

Restaurants

Regional distributors

22,600 farms$1.2 billion in sales

CSA=Community-supported agriculture.Source: USDA, Economic Research Service calculations based on 2008 Agricultural Resource Management Survey, conducted byUSDA, National Agricultural Statistics Service and Economic Research Service.

ExclusivelyExclusi

vely


10/38

4



Which Marketing Channels Do Local FoodSales Farms Use?

Given that the market or local oods is more extensive than previouslymeasured, we explored how local ood sales arms (by arm sales class)

dier in their use o marketing options. According to the 2008 ARMS,small local ood arms (gross arm sales less than $50,000) representedalmost 81 percent o all local ood arms; medium-sized arms (gross armsales $50,000-$249,999) represented 14 percent; and large arms (sales o$250,000 or more) accounted or almost 5 percent o all local ood arms(table 1).

The ratio o local ood sales to total arm sales measures how these armsrely on local ood sales or their nancial viability.5 For all sizes o local oodarms, marketing local ood products accounts or 61 percent o gross armsales, on average (table 1). The median ratio o local ood sales to total armsales is higher than the mean ratio or each arm size, highlighting the act

that many o these arms rely on local ood sales. Almost two-thirds o alllocal ood producers reported that local ood sales accounted or at least 75percent o their total gross arm sales, while 22 percent o all local ood salesarms reported that such sales accounted or less than 25 percent o their totalgross arm sales. Higher local ood sales shares suggest that local ood salesarms are well integrated into existing direct-to-consumer and intermediatedsupply chains.6

A dierence exists in the size class distribution o direct-to-consumer sales,as observed in the 2007 Census o Agriculture, and total local oods sales, asobserved in the 2008 ARMS. Small arms in 2008 accounted or 11 percento total local ood sales; medium-sized arms accounted or 19 percent o

total local ood sales; and large arms accounted or almost 70 percent (table1). These shares o total local ood sales based on arm size are at odds withthose or direct-to-consumer sales only, in which each sales class accountedor roughly a third o direct-to-consumer sales.

Small arms are more likely to market through direct-to-consumer outlets,perhaps because small arms cannot generate enough volume or distributorsand institutions that demand high volumes o local ood (i.e., grocers, restau-rants, and schools) (Gale, 1997; Brown, 2002). Small arms account or 81percent o all local ood sales arms and many rely on exclusive use o direct-

5About 20 percent o all arms are point

sales arms (sales less than $1,000).

Due to statistical reliability problems

associated with ew observations, this

study was not able to report any local

ood arms classied as point sales

arms. This suggests that point saleslocal ood arms may have been part

o the 146,000 armers excluded rom

this study that reported direct sales

to consumers but ailed to report the

marketing channels used or the value

o their direct sales (see Appendix 1 or

urther discussion).

6Local oods specialists and analysts

studying the economic viability o small-

and medium-sized arms advocate

public- and/or private-sector initiatives

to deepen and widen the inrastructure

o these supply chains (Clancy andRuh, 2010; Matteson and Heuer, 2008).

As an example o a private-sector

initiative on local ood, Walmarts

Heritage Agriculture program has been

covered in the press including The New

York Times (Cliord, 2010) and The

Atlantic Monthly (Kummer, 2010),

but no study to date has looked at this

programs eects on small arms and

local communities.

Comparing ARMS Data on Local Foods Sales

Prior to the 2008 ARMS, data collected on local ood sales to consumers re-

erred to direct sales o agricultural products or human consumption in the

Census o Agriculture and ARMS surveys. The 2008 ARMS questionnaire

asked an extensive set o questions on armers direct sales; however, the

phrase or human consumption was omitted. As a result, the 2008 ARMS

data had to be manipulated to provide estimates o direct ood sales that were

consistent with estimates rom previous ARMS and census o agriculture data.*

* See Appendix 1: Developing Data on Marketing Local Agricultural Products

Using the 2008 ARMS or a discussion o the 2008 ARMS questionnaire and how

sales o local oods were estimated or this study.


11/38

5


to-consumer marketing channels, grossing $6,740 in sales on average in 2008(table 1). The share o small- and medium-sized arms exclusively marketingthrough direct-to-consumer channels is higher than the corresponding shareo large arms.

Selling local oods through intermediated outlets may require less arm laborthan selling via direct-to-consumer outlets because armers are not requiredto spend time at intermediated outlets (Brown, 2002). Larger arms may have

a comparative advantage in intermediated sales because many restaurants,grocers, and regional distributors demand timely delivery o large volumes oood with consistent quality. In 2008, large local ood sales arms accounted

Table 1

Marketing channels used by local food sales farms, by farm size

Farm size

IItem

Small(sales o less than

$50,000)

Medium(sales o $50,000-

$249,999)

Large(sales o $250,000

or more) All

----------------------------------Number----------------------------------Local ood sales arms 86,726 15,202 5,301 107,229

----------------------------------Percent-----------------------------------

Local ood sales arms 80.9 14.2 4.9 100.0

All arms 5.3 5.1 2.5 5.0

Average ratio o local ood sales to totalarm sales 68.8 67.2 57.5 61.2

Median ratio o local ood sales to totalarm sales 100.0 80.0 80.0 100.0

Farms by marketing channels 100.0 100.0 100.0 100.0

Direct-to-consumer channels only 72.1 46.5 31.0 66.4

Intermediated marketing channels only 11.3 10.4 37.1 12.5

Both marketing channels 16.6 43.0 31.9 21.1

Local ood sales: ---------------------Percent---------------------- Million dollars

Marketed through all channels 11.1 19.1 69.8 4,806

Direct-to-consumer channels only 33.7 38.9 27.4 887

Intermediated marketing channels only 3.5 3.6 92.9 2,720

Both marketing channels 11.7 39.5 48.8 1,199

Average local ood sales per arm: -----------------------------------Dollars-----------------------------------

Marketed through all channels 7,856 69,985 771,965 56,240

Direct-to-consumer channels only 6,737 66,247 305,181 17,621

Intermediated marketing channels only 10,242 73,126 1,338,257 217,150

Both marketing channels 9,768 72,312 352,375 53,103

Source: USDA, Economic Research Service calculations based on 2008 Agricultural Resource Management Survey, conducted byUSDA, National Agricultural Statistics Service and Economic Research Service.


12/38

6



or 93 percent o the $2.7 billion in sales generated exclusively through inter-mediated channels, averaging $1.3 million in local ood sales per arm (seetable 1). Medium and large local ood sales arms together accounted or 88percent o almost $1.2 billion in sales marketed by arms using both interme-diated and direct-to-consumer outlets.

Small local ood sales arms gross, on average, $10,240 per arm annuallywhen marketing exclusively through intermediated channels (see table 1).

Formal and inormal collaboration with other armers provides a way orthese small arms to meet the quantity, quality, packaging, and deliveryrequirements o grocers and restaurants (PFI, 2009). Medium-sized armsaccounted or 17 percent o ood arms relying solely on intermediatedmarketing channels, averaging $203,900 in local ood sales per ood arm.

In 2008, 107,000 local ood sales arms reported using 160,800 marketingchannels to sell local ood (table 2). Direct-to-consumer outlets accountedor approximately 75 percent o these marketing channels. Roadside standsand armers markets accounted or about 80 percent o the direct-to-consumer outlets used by armers. According to the 2008 ARMS, armersselling local ood at armers markets traveled an average 30.7 miles, drivingpast the nearest town o 10,000 residents to their destination, suggestingthat small towns may not generate enough consumer demand to supportarmers markets.7 Onarm stores and CSAs were used much less requently.Intermediated outlets accounted or the remaining 25 percent o local oodmarketing channels used by armers.

7 Farmers travel 30.7 miles to their

armers market, on average, whereas

their nearest town o 10,000 or more

residents lies 5.3 miles away. The

median distance traveled to a armers

market was 15 miles, while the maxi-

mum distance was 275 miles.

Table 2

Local food marketing channels used, by farm size

Farm size

Sales channels

Small(sales o less than

$50,000)

Medium(sales o $50,000-

$249,999)

Large(sales o $250,000

or more) All

Number

Local ood sales outlets used 121,198 15,202 5,301 160,795

Average number o outletsused per arm 1.4 1.7 2.1 1.5

Percent

By marketing outlet 100.0 100.0 100.0 100.0

Direct-to-consumer outlets 78.0 70.7 55.5 75.3

Roadside stands 34.1 24.9 23.7 31.8

Farmers markets 34.6 25.9 14.7 31.8

Onarm stores 8.3 17.4 15.7 10.4

CSAs 1.1 2.5 1.4 1.3

Intermediated outlets 22.0 29.3 45.0 24.7

Grocers and restaurants 17.2 26.0 23.7 19.2

Regional distributors 4.8 3.4 21.4 5.5

CSAs=Community-supported agriculture.Source: USDA, Economic Research Service calculations based on 2008 Agricultural Resource Management Survey, conducted byUSDA, National Agricultural Statistics Service and Economic Research Service.


13/38

7


As the size o local ood sales arms increases, the requency o arms sellingthrough direct-to-consumer marketing channels declines and the requency osales through intermediated marketing channels increases. Small local oodarms are three times more likely to use direct-to-consumer outlets than inter-mediated outlets (see table 2). With larger sales, large local ood sales armsdivide local ood sales in a 55-45 split between direct-to-consumer and inter-mediated marketing channels. Reducing direct-to-consumer marketing likelyreduces marketing costs or these large arms.

Consumers may acquaint the public ace o local oods with armers usingdirect-to-consumer outlets because they represent most producer-consumerinteractions (see table 2). Small- and medium-sized arms account or 95percent o direct-to-consumer local ood sales arms. In a variety o surveys,consumers reported that consumer-armer interactions and consumersdesires to support local producers were as important as the quality o thecommodity (Hunt, 2005; Brown and Miller, 2008; Thilmany et al., 2008).

Direct-to-consumer marketing channels, however, are not how most localoods are purchased; at least 60 percent o the value o local ood salespassed through intermediated channels dominated by large ood arms(see table 1). Could consumer interactions with small and medium armersat direct-to-consumer outlets have translated into increased local oodspurchases at grocery stores and restaurants? The popular press assumes thisto be the case, but the extent to which consumer-armer interactions at direct-to-consumer outlets have infuenced retail purchases o local oods has yet tobe tested empirically.

What Commodities Are Being Producedfor Local Food Sales?

Just as marketing outlets vary, commodities produced by local ood armsdier rom all U.S. arms. Local ood arms principally produce resh vege-tables, ruits, and nuts, contrasting with traditional arm production, which isprincipally composed o livestock and program commodity crop production.According to the 2008 ARMS, vegetable, ruit, and nut arms representedalmost 6 percent o the 2.1 million arms, yet they accounted or 43 percento all local ood arms and generated $3.0 billion, or 65 percent, o total saleso locally grown ood.8 While only 5 percent o all arms engaged in localood sales, about 40 percent o vegetable, ruit, and nut arms sold throughlocal ood channels. That is, a vegetable/ruit/nut arm is eight times morelikely to sell ood commodities locally than other arms.

Vegetable, ruit, and nut arms also rely more on local ood sales to generate

gross arm sales than eld crop or livestock arms. Among all local oodarms, local ood sales account or 65 percent o gross arm sales or ruit,vegetable, and nut arms, on average, but only 37 percent or livestock andeld crop arms. Excluding local ood sales arms marketing solely throughintermediated channels, vegetable, ruit, and nut arms grossed $32,000 perarm in local ood sales in 2008 compared with $13,800 per arm or eldcrops and livestock arms.9

Vegetable, ruit, and nut arms participate at varying levels in the threemajor marketing channel combinations, but they account or the largest share

8ARMS classies arms into 19 produc-

tion types according to the agricultural

commodity that accounts or at least

50 percent o arm sales. For more

inormation, see USDA, NASS, 2009.

For the purposes o this study, we ag-

gregated all ood arms into three basic

categories: ruit/vegetable/nut arms, all

other eld crop arms, and arms pro-ducing livestock and livestock products.

9Including armers who use only

intermediated channels would skew the

distribution o average sales per arm

even urther. Among those local ood

sales arms that rely solely on interme-

diated marketing channels, vegetable,

ruit, and nut arms grossed $509,400

per arm compared with $105,900 per

arm or eld crops and livestock arms.


14/38

8



o sales in each combination. Vegetable, ruit, and nut arms represent 40percent o armers using direct-to-consumer sales exclusively, 25 percento armers using intermediated channels exclusively, and almost 60 percento armers using both types o marketing channels. Vegetable, ruit, and nutarms generate roughly 60 percent o all local ood sales that pass throughdirect-to-consumer and intermediated channels, and over 70 percent o localood sales marketed by arms using both types o channels.

The disproportionate presence o vegetable, ruit, and nut arms among alllocal ood arms shapes the typical prole o local ood sales arms. Thesearms operate ewer acres while generating higher gross sales per acre thaneld crop or livestock arms.10 The average local ood sales armer growshigh-valued ood commodities on 149 acres that yield, on average, $590 peracre in sales. In contrast, the operator o the average arm generates $304 insales per acre on 392 acres.

Comparing Farms That Market Local FoodsWith Farms That Do Not Market Local Foods

By some measures, a higher percentage o armers who market local oods

appear to devote more time to arming as an occupation than is the case orarmers who do not market local oods. In particular:

Primary operators o local ood sales arms are 30 percent more likely tolist their primary occupation as arming (table 3). Small local ood sales

10Vegetable, ruit, and nut arms, on

average, generate $1,338 per acre in

sales on 76 acresour to six times the

revenue per acre on a arm that is 33-50

percent the size o the average eld

crop or livestock arm. Average gross

sales per acre ranges rom $640 per

acre or vegetable, ruit, and nut arm-

ers using direct-to-consumer outlets

only to $1,310 per acre or those using

both direct-to-consumer and interme-diated outlets, and to over $3,100 per

acre or those relying exclusively on

intermediated outlets.

Table 3

Farms that have local food sales compared with those with nolocal food sales

ItemFarms with

local ood salesFarms with no

local ood sales

Primary operator characteristics:

Age o the primary operator 57.2 57.8

Women as primary operators(percent o all arms) 10.2 10.5

Beginning armers (with 10 years or lessexperiencepercent o all arms) 25.4 23.3

Age frst became arm operator* 33.7 31.7

Years o experience as an operator ** 23.4 25.9

Years o education*** 14.0 13.2

Internet use (percent o arms)* 69.9 63.4

Measures o operator commitment to arming:

Farming as primary occupation(percent o arms)*** 58.3 44.6

Full-time equivalent operator jobs per arm*** 1.3 0.9

Either one or both spouses work in o-armjobs (percent o arms) 57.3 61.4

Average o-arm labor income (dollars)* 36,739 44,196

Note: Dierence-o-means test statistics (t) were calculated or each variable. *Statisticallysignifcant at the 10-percent level. **Statistically signifcant at the 5-percent level. ***Statisticallysignifcant at the 1-percent level.Source: USDA, Economic Research Service calculations based on data rom the 2008Agricultural Resource Management Survey, conducted by USDA, National AgriculturalStatistics Service and Economic Research Service.


15/38

9


arm operators are 50 percent more likely to do so, but this dierencedisappears or large local ood sales arm operators.

Household members o arms marketing local oods devote more time toarm operation than do household members o arms that do not marketlocal oods. Local ood sales arms devote 40 percent more operator worktime to arminglling 1.3 operator ull-time-equivalent jobs (1 FTEequals 2,000 hours worked annually) compared with a 0.9 operator FTE

job or the average arm.

Farm households that sell local oods earn 17 percent less, on average,in o-arm labor income than average arm households that do not selllocal oods.

These measures suggest that the occupational and time commitments toarming are valued more by local ood sales arm households than the ore-gone labor income they could have earned o arm.

To examine how local ood armers commitment may translate intoincreased arm business viability, we compared two nancial perormance

measures between local ood arms and arms without local ood sales:

Farms earning positive prots.

Mean operating expense ratios.

The same share o arms with and without local ood sales earned positiveprots. For the lowest and highest sales classes, we ound some statisticalevidence o dierences in mean operating expense ratios (dened as totalcash expenses divided by gross cash income); however, they were notdetected or the sample as a whole. Once armers pass $10,000 in annualgross sales, operating expense ratios o arms engaged in local ood salesmay be lower than the average arm not engaged in local ood sales, implying

that local ood sales arms may reach protability at a lower gross sales point.

When comparing other arm operator characteristics, we detected dier-ences in experience and education between arm operators who market localoods and those who do not market local oods. Operators o local ood salesarms have an average o 2 years less experience in arming, and they startedarming 2 years later in lie than the average armer (see table 3). Local oodsales arm operators have completed an average o 1 more year o educationand are about 6.5 percent more likely to use the Internet. We did not ndsignicant dierences in operator characteristics with regards to gender, theaverage age o the primary operator, beginning armers as a percentage o allarmers, or whether one or both spouses worked in o-arm jobs.


16/38

10



Location of Local Food Sales Farms

To better understand where local ood sales arms are located, we supple-mented our analysis o the 2008 ARMS data with direct-to-consumer salesdata rom the 2007 Census o Agriculture. Modeling direct-to-consumersales with Census o Agriculture data allows us to examine the location oproduction, while controlling or other actors (e.g., urbanization and crop-

land availability).11 There were not enough respondents with local ood salesin the 2008 ARMS sample to model local ood production.

We ound that proximity to a metro area, access to armers markets andarmland, and being in the coastal regions o the United States are driverso direct-to-consumer sales, but we cannot say much about consumer-sidedrivers o direct-to-consumer sales (e.g., demographics). Our results suggestthat local ood sales have the greatest potential or economic developmentin specic places and regions o the country. These results are consistentwith prior research also using the 2007 Census o Agriculture (Vogel andLow, 2010).

Local Food Sales in U.S. Regions

Local ood sales vary regionally. Direct-to-consumer sales are highest inthe Northeast, on the West Coast, and around a ew isolated metropolitanareas throughout the country (g. 3). Even ater controlling or urbaniza-tion, our analysis o direct-to-consumer sales suggests that such sales aresignicantly higher on the West Coast and in the Northeast.12 This resultcorrelates with the evidence on both direct-to-consumer sales and intermedi-ated sales rom 2008 ARMS data and other research (USDA-AMS, 2009).Direct-to-consumer sales analysis suggests that some actors aecting thesupply o direct-to-consumer sales are infuenced by neighbors, while othersare regionally infuenced. For example, arms with direct-to-consumer sales

are most likely to have neighbors who also participate in direct salesthis isa neighborhoodeect rather than a regional eect. Direct-to-consumer salesare highest in regions that produce more ruits and vegetablesa result likelydriven by the geographic suitability or growing ruits and vegetables (e.g.,regional climate, topography, and inrastructure).

According to 2008 ARMS data, arms on the West Coast (Caliornia,Oregon, and Washington State) with local ood sales accounted or only 7.8percent o all local ood sales arms, but they accounted or 23.8 percento all local ood sales and 31.4 percent o all local oods sales o ruit, nut,and vegetable sales. Recognized or its varied microclimates, long growingseason, and extensive irrigation networks, the West Coast supplies the Nation

with 56 percent o all vegetables, ruits, nuts, and other specialty crops.

The West Coast has a long-standing system o armers markets and armer-to-grocers marketing channels dating back to the 1970s. Small-scale armersbegan selling organic and high value-added niche oods to upscale restau-rants in the late 1970s (now a national trend) and are now part o arm-to-school marketing arrangements. Another U.S. hot spot or local ood salesis the Atlantic seaboard, particularly the Northeast census division. Localood sales arms in the Northeast generated 14.4 percent o U.S. local ood

11We estimated county-level andcommuting-zone level spatial econo-

metric models o actors correlated

with direct-to-consumer sales and its

location using 2007 Census o Agricul-

ture data on direct sales as a dependent

variable. For a detailed discussion

about the econometric analysis, see

Appendix 2.

12This result would likely be even

stronger i intermediated sales were

included in Census o Agriculture data.


17/38

11


production. Regional dierences in local ood sales may be explained by theavailability o logistics and distribution inrastructure (e.g., King et al., 2010).

Direct-to-consumer and intermediated ood sales marketing practices dierbetween regions. West Coast local ood sales arms are more likely thanthose in the Northeast to be large arms located arther rom metro areas. Asa result, these arms predominantly market through intermediated marketingoutlets, which are less time and eort intensive than direct-to-consumer

marketing outlets. Indeed, 85 percent o West Coast local ood sales occurredthrough intermediated channels. Local ood sales arms in the Northeast tendto be smaller, located closer to densely populated urban markets, and morelikely to use only direct-to-consumer marketing outlets.

The share o local ood producers who are beginning armers with 10 yearso experience or less also varies regionally and is highest in the West. Forty-eight percent o West Coast local ood producers are beginning armers as are28 percent o Northeast local ood producers, both higher than the nationalshare o 24.3 percent. More beginning armers may be driven by high localood demand, but without data on the same producer over time, it is dicultto understand why more beginning armers are located on the West Coast and

in the Northeast.

Figure 3

Value of direct-to-consumer sales, by county, 2007

Median sales or less

$123,000 up to $1 million

$1 million up to $2.5 million

$2.5 million or more

Not available/disclosure issues

Legend

Source: USDA, National Agricultural Statistics Service; 2007 Census of Agriculture.


18/38

12



Local Food Sales Highest in Urban Areas

According to the 2008 ARMS, over hal o all arms with local ood saleswere located in metropolitan counties, compared with only a third o allU.S. arms.13 Using Census o Agriculture data on direct-to-consumer sales,we ound that even when controlling or region-specic actors, the domi-nance o direct-sales arms in metropolitan counties holds (see Appendix2: Modeling Direct-to-Consumer Sales), suggesting that proximity to urbanmarkets is strongly related to the production o directly sold goods. Therewere not enough observations in the ARMS data to test metropolitan domi-nance, but tabular evidence suggests this result would not change i interme-diated sales were included.

The dominant metro location o direct-sale arms could be driven bydemand-side actors (e.g., access to thickly populated markets and armersmarkets) or by supply-side actors (e.g., access to labor, agricultural land,or transportation networks). Our analysis suggests that, all other actorsbeing equal, both demand-side actors and supply-side actors, includingregional production o ruits and vegetables and availability o tillable land,aect direct-to-consumer sales at the county and commuting-zone level. Ourevidence did not nd a clear correlation between consumer characteristics(population demographics) and direct-to-consumer sales, however.

More than 50 percent o small local ood sales arms were ound in metrocounties and 30 percent in rural counties adjacent to metro counties, whilenonlocal ood sales arms were, on average, more equally distributed acrossmetro, adjacent rural, and remote rural counties. On the demand-side, metro-politan concentration gives producers access to the urban local ood salesmarkets essential to their economic viability. On the supply-side, concen-tration near urbanized areas may be the result o urban development pres-sures on land prices leading to the dissolution o large arms. As such, the

remaining arm operations are smaller and must produce higher valued, nicheagricultural commodities (Heimlich and Anderson, 2001; Nickerson, 2001).

Larger local ood arms are more likely to be located in remote, nonmetro-politan areas. ARMS estimates show that 50.1 percent o large arms withlocal ood sales were located in nonmetro counties not adjacent to metropol-itan areasonly 32.6 percent o all large arms were located in these remotecountiesand there were ar ewer o these large local ood arms.

Our analysis o direct-to-consumer sales data rom the Census o Agriculturesuggests that, while controlling or urbanization, the availability o arm-land and other costs may drive location decisions o arms with direct-to-

consumer sales. The availability, cost, and quality o labor may aect localood sales arms because they are more labor-intensive than comparablearms not engaging in direct sales. Fruit and vegetable arms with local oodsales employed 61,000 workers in 2008, or 13 FTE employees per milliondollars o sales, while ruit and vegetable arms not engaged in local oodsales employed only 3 FTE employees per million dollars o sales.

13We used the Oce o Manage-

ment and Budgets 2003 deni-

tion o metropolitan counties in

this analysis. Counties outside but

adjacent to a metropolitan statisti-cal area are reerred to as adjacent

counties. Remote counties are those

outside o and not adjacent to a

metropolitan area.


19/38

13


Conclusion

The 2008 ARMS data provide a broader picture o arms engaged inmarketing local oods. By assessing both direct-to-consumer and intermedi-ated sales, we can develop a more complete picture o local ood marketsand producers. Local ood sales via intermediated marketing channels are animportant component o the industry that has not previously been extensively

studied.

We ound that small arms with gross sales under $50,000 accounted or81 percent o local ood sales arms and were more likely to use direct-to-consumer marketing channels, such as armers markets and roadside stands,exclusively. Making up 14 percent o all local ood sales arms, medium-sized arms were equally likely to use only direct-to-consumer marketingchannels or a mixture o direct-to-consumer and intermediated marketingchannels, with only 11 percent using intermediated channels exclusively.Combining marketing channels may represent the appropriate marketstrategy or medium-sized arms to thrive. Large arms represented 5 percento all local ood sales arms. Most local ood sales by large arms were

marketed by those exclusively using intermediated channels. In doing so,these arms were able to reduce labor expenses per dollar o sales by leavingthe labor-intensive distribution o local oods up to intermediaries.

According to the 2008 ARMS, or small and medium local ood sales arms,more primary operators identied their primary occupation as arming and alloperators devoted more work time to production than similarly sized armswithout local sales.

Our model o the location o producers with direct-to-consumer sales andthe analysis o direct-to-consumer and intermediated local ood sales withrespect to location indicates that local ood sales are a regional phenomenon

and that marketing practices vary among regions. Controlling or variousproduction actors, direct-to-consumer sales were highest in and near urbanareas and production likely depended on the availability o labor, tillableland, and the market inrastructure essential or direct-to-consumer sales.Policy decisions that oster local ood sales must account or the importanceo vital, but unalterable, regional characteristics, such as climate, water avail-ability, and access to densely populated markets, which aect the viabilityo local oods as an economic development tool. Findings suggest that localood sales have the potential or community economic development incertain areas o the country, particularly those close to urban areas.

While the ndings o this study provide additional quantitative inormationat the national level on armers engaged in local ood marketing, the 2008ARMS data are not without problems. These data are not comparable withprevious USDA direct sales estimates. Additionally, we do not have dynamicdata that might enable us to understand the tenure and success o local oodsales arms. Further work is necessary to understand the protability o localood sales arms and noneconomic reasons or direct-to-consumer marketing.

Improving data collection methods on local oods occurs iteratively giventhe time span between developing and rening the current years ARMS


20/38

14



questionnaire. The 2009 ARMS questionnaire restored the phrase or humanconsumption. The 2010 ARMS separates the value o direct-to-consumersales rom intermediated marketing sales and, or the rst time, also includesinstitutional sales as part o the intermediated marketing category. These datasets are available to researchers.


21/38

15


References

Anselin, Luc. Spatial Econometrics: Methods and Models (Studies inOperational Regional Science), Boston: Kluwer Academic Publishers,1988.

Brown, Allison. Farmers market research 1940-2000: An inventory and

review,American Journal o Alternative Agriculture 17(4), pp. 167-76,2002.

Brown, Cheryl, Jesse E. Gandee, and Gerard DSouza. West Virginia FarmDirect Marketing: A County Level Analysis,Journal o Agricultural andApplied Economics 38(3), pp. 575-84, 2006.

Brown, Cheryl, Stacy Miller, Deborah Boone, Harry Boone, Stacy Gartin,and Thomas McConnell. The importance o armers markets or WestVirginia direct marketers,Renewable Agriculture and Food Systems22(1), pp. 20-9, 2007.

Brown, Cheryl, and Stacy Miller. The Impacts o Local Markets: A Reviewo Research on Farmers Markets and Community Supported Agriculture(CSA),American Journal o Agricultural Economics 90(5), pp. 1296-1302, 2008.

Clancy, Kate, and Kathryn Ruh. Is Local Enough? Some Arguments orRegional Food Systems, Choices Magazine, Agricultural and AppliedEconomics Association, 25(1), 2010.

Cliord, Stephanie. Wal-Mart to Buy More Local Produce, The New YorkTimes, October 14, 2010.

Gale, Fred. Direct Farm Marketing as a Rural Development Tool,RuralDevelopment Perspectives , U.S. Department o Agriculture, EconomicResearch Service, 12(2), pp. 19-25, 1997.

Hand, Michael, and Stephen Martinez. Just What Does Local Mean?Choices Magazine, Agricultural and Applied Economics Association,25(1), 2010.

Heimlich, Ralph E., and William D. Anderson.Development at the UrbanFringe and Beyond: Impacts on Agriculture and Rural Land, AER-803,U.S. Department o Agriculture, Economic Research Service, p. 88, 2001,http://www.ers.usda.gov/publications/aer803/.

Hoppe, Robert, Penni Korb, and David E. Banker.Million-Dollar Farms inthe New Century, EIB-42, U.S. Department o Agriculture, EconomicResearch Service, p. 41, 2008, http://www.ers.usda.gov/publications/eib42/.

Hunt, Alan. Conserving the Agricultural Landscape Through FarmersMarkets, masters thesis, Duke University, p. 130, 2005.


22/38

16



King, Robert P., Michael S. Hand, Gigi DiGiacomo, Kate Clancy,Miguel I. Gomez, Shermain D. Hardesty, Larry Lev, and Edward W.McLaughlin. Comparing the Structure, Size, and Perormance o Localand Mainstream Food Supply Chains, ERR-99, U.S. Department oAgriculture, Economic Research Service, June 2010.

Kummer, Corby. The Great Grocery Smackdown: Will Walmart, not WholeFoods, save the small arm and make America healthy? The Atlantic

Monthly, March 2010.

Lev, Larry, and Lauren Gwin. Filling in the Gaps: Eight Things toRecognize about Farm-Direct Marketing, Choices Magazine,Agricultural and Applied Economics Association, 25(1), 2010.

Lyson, Thomas, and Amy Guptill. Commodity Agriculture, CivicAgriculture, and the Future o U.S. Farming,Rural Sociology 69, pp.370-85, 2004.

Martinez, Steve, Michael Hand, Michelle DaPra, Susan Pollack, KatherineRalston, Travis Smith, Stephen Vogel, Shellye Clark, Luanne Lohr, Sarah

Low, and Constance Newman.Local Food Systems: Concepts, Impacts,and Issues, ERR-97, U.S. Department o Agriculture, Economic ResearchService, May 2010.

Matteson, Gary, and Robert Heuer. Growing Opportunity: Outlook or theLocal Food Systems Marketplace, unpublished manuscript, Washington,DC: Farm Credit Council, Young, Beginning, and Small Farmer Program,p. 39, 2008.

Morgan, Tamekia, and Dovi Alipoe. Factors Aecting the Number andType o Small-Farm Direct Marketing Outlets in Mississippi,Journal oFood Distribution Research 32, pp.125-32, 2001.

Nickerson, Cynthia. Smart Growth: Implications or Agriculture in UrbanFringe Areas,Agricultural Outlook/April 2001 , pp 24-7, 2001,http://www.ers.usda.gov/publications/agoutlook/april2001/AO280g.pd.

Practical Farmers o Iowa (PFI). To Market, To Market: Practical Farmerso Iowa Field Day Explores Grower Collaboration, 2009, www.practi-calarmers.org/news.

Thilmany, Dawn, Craig A. Bond, and Jennier K. Bond. Going Local:Exploring Consumer Behavior and Motivations or Direct FoodPurchases,American Journal o Agricultural Economics 90(5): pp. 1303-09, 2008.

Thilmany, Dawn, and Philip Watson. The increasing role o directmarketing and armers markets or Western U.S. producers, WesternEconomics Forum, pp. 19-25, April 2004.

Tolbert, C., and Molly Sizer. U.S. Commuting Zones and Labor MarketAreas: A 1990 Update, AGES-9614, U.S. Department o Agriculture,Economic Research Service, 1996.


23/38

17


U.S. Department o Agriculture, Agricultural Marketing Service. USDANational Farmers Market Manager Survey 2006, 2009.

U.S. Department o Agriculture, Economic Research Service. 2007 and 2008Agricultural Resource Management Surveys (ARMS).

U.S. Department o Agriculture, Economic Research Service. 2000 Rural-Urban Commuting Area Codes, 2005, http://www.ers.usda.gov/brieng/

Rurality/RuralUrbanCommutingAreas/.

U.S. Department o Agriculture, National Agricultural Statistic Service.2007 Census o Agriculture, 2009.

U.S. Department o Agriculture, National Agricultural Statistics Service.2008Agricultural Resource Management Survey (ARMS) Phase II Costand Returns Report, Project 904 - ARMS Phase III CRR Version1 QID 044-001, p. 36, 2008, http://www.ers.usda.gov/Data/arms/app/ARMSDocs/Questionnaires/W%5E2008%5ECRR%5EPhase3%20Questionnaire%5EQ%5EFOH.pd.

U.S. Department o Agriculture, National Agricultural StatisticsService. Section I Farm Management and Use o Time, 2008Agricultural Resource Management Survey (ARMS) Phase II

Cost and Returns Report: Interviewers Manual, p. 391, January2009, http://www.ers.usda.gov/Data/ARMS/app/ARMSDocs/Questionnaires/W%5E2008%5EAll%5EPhase3%20Interviewers%20Manual%5EM%5ECOP_FOH.pd.

Vogel, Stephen, and Sarah A. Low. Farmers direct sales to consumersgrowing,Amber Waves, U.S. Department o Agriculture, EconomicResearch Service, September 2010.


24/38

18



Appendix 1Developing Data on MarketingLocal Agricultural Products Using the2008 ARMS

The 2008 ARMS is the rst nationally representative survey to queryarmers about the local marketing channels they used to sell their agriculturalcommodities. The intent o the questions in the 2008 ARMS questionnairewas to ocus on marketing local oods (USDA/NASS, 2009). However, thedesign and structure o the questions created obstacles to deriving estimatesthat were both internally and externally consistent. Internal consistencymeans that tabulated responses to one or more questions must be consis-tent with tabulated responses to subsequent questions. External consistencymeans that the 2008 ARMS must produce estimates on direct-to-consumersales o ood as close as possible to estimates generated rom the 2007ARMS and 2007 Census o Agriculture when attempting to measure thesame phenomenon.

Results or questions 21-28 o Section I, Farm Management and Use o

Time, o the 2008 ARMS questionnaire are presented in appendix table 1.To see the exact wording o each question, reer to the end o this appendix.I a respondent answered yes to questions 21a or 21b, the respondentproceeded to the remaining questions in the module. Question 22 asks thearmer a set o questions related to the commodities sold and i processingwas required beore sale. Question 23 queries which marketing channels thearmer used to sell a product. Question 23 allows us to distinguish betweendirect-to-consumer and intermediated sales. Question 24 asks what share ototal arm sales were marketed through any o the channels listed in question23, which allows us to calculate the value o local sales. Questions 25-28asks the armer about other practices indirectly linked to sales marketed inthe channels listed in question 23.

We used the armers positive responses to questions 23a-23e and 23g toconstruct an estimate o the number o armers selling arm goods locallyto consumers. Responses to question 23 on State branding o arm productswere excluded because State-branded products are oten marketed nationally,perhaps even internationally. According to appendix table 1, the number oarms using one or more marketing channels accounted or 134,200 armersmarketing $8.0 billion o ood and nonood products through 199,000 direct-to-consumer (23a-23d) and intermediated outlets (23e, 23g).

Since question 21a does not include the phrase or human consumption orspeciy i local channels were used, positive responses to this question gener-

ated an estimated 280,100 arms selling arm goods directly to consumersmore than twice the number o armers who reported direct sales or humanconsumption in the 2007 ARMS and the 2007 Census o Agriculture. Thismeans that 145,900 armers reported selling arm output to consumers in2008 but ailed to speciy their marketing channels and the volume o directsales. Two possibilities may explain this result. First, responses to question21a may capture sales o ood and nonood products without distinguishingbetween local and national direct marketing channels, such as Internet ormail-order sales. Second, responses to this question may capture point salesto consumers by armers whose sales were too low to report using direct


25/38

19


marketing supply chains. These armers may be just testing the waters.

Hence, this question by itsel adds no useul inormation and was omittedrom our evaluation. Question 21b also suers rom this same lack omarketing-channel and geographic specicity and was not used.

The 2008 ARMS survey does not allow the researcher to separate out thevalue o sales o ood and nonood products sold directly by a respondent,orcing us to rely on indirect methods to estimate the value o ood sales.We used the Economic Research Service (ERS) arm production typologyto separate out direct sales arms selling ood products and those sellingnonood products. The ERS arm typology categorizes arms by commodityi a particular commodity accounts or at least 50 percent o arm sales.Farms categorized as Nursery and Cut Tree Farms were classied as

nonood arms and accounted or over 95 percent o direct nonood sales.1Excluding this category generated 107,200 estimated amers marketing $4.8billion in local ood products through 160,000 channels (see appendix table 1).

As a result, 2008 ARMS estimates o armers engaging in local ood salesappear similar to estimates rom the 2007 ARMS and 2007 Census oAgriculture. That the estimated value o local ood sales is our times higherthan previous estimates suggests that new inormation in the 2008 ARMSprovides a broader, more complex picture o armers engaged in directmarketing o local oods.

The 2008 ARMS did not collect data linking the value o sales to the use o

a particular marketing channel. Question 24asked or the volume o salesassociated with any o the channels listed in question 23. The design o thisquestion orced us to adhere to a strict trichotomy, grouping sales by exclu-sive use o direct-to-consumer outlets, exclusive use o intermediated chan-nels, or marketing simultaneously through both channels (see appendix table1). I we were to try to tease out the value o local ood sales by marketingchannel, we would encounter problems with double counting, condenti-ality, and statistical reliability. For those arms using both types o marketing

1Although nurseries and cut tree armscan and do sell ood items directly

to consumers, we have no way o

quantiying their ood sales separate

rom direct sales o nonood products

in the 2008 ARMS. According to the

2007 ARMS, 3,400 nurseries and cut

tree arms, or 7 percent o all nursery

and cut tree arms, sold $171 million

o ood products directly to consum-

ers. For this small segment o nursery

and cut tree arms, direct sales o local

ood are important to the viability o

their arm operations, accounting or

34 percent o total gross sales. For thesector as a whole, however, direct sales

o ood represent less than 2 percent

o total gross sales o nursery products

and cut trees. Less than 2,000 arms

sold directly to consumers and were

also categorized as primarily producing

nonood products, such as horses, other

live animals not or meat consumption,

and aquaculture. These arms ac-

counted or very little o the direct sales

reported in question 23.

Appendix table 1

Direct sales farms, by farm size, 2007 Census and 2007 and 2008 ARMS

Data sourceNumber o direct

sale arms Percent o all armsGross value

o direct salesPercent o all

arm sales

Thousands Percent

Millions

(nominal dollars) Percent

2007 Census o Agriculture 136.8 6.2 1,211 0.4

2007 ARMS 115.5 5.3 1,292 0.6

2008 ARMS all arms 134.2 6.3 8,058 3.2

2008 ARMS local ood sales arms 107.2 5.0 4,806 1.9

Direct-to-consumer 71.2 3.3 877 0.4

Intermediated 13.4 0.6 2,715 1.0

Both marketing channels 22.6 1.1 1,198 0.5

Source: USDA, National Agricultural Statistics Service 2007 Census o Agriculture; 2007 and 2008 Agricultural ResourceManagement Survey, Version 1, conducted by the National Agricultural Statistics Service and Economic Research Service.


26/38

20



channels, the data did not allow us to quantiy the contribution each type omarketing channel makes to overall arm perormance.

Optimal marketing strategies are o particular importance when assessing theviability o the medium-sized arms that make the greatest use o both typeso channels. Clancy and Ruh (2010) pointed out that without direct andintermediated marketing channels, sucient scale economies or mid-sizedarms cannot be achieved.

The data presented additional challenges or our study, namely our abilityto distinguish how local ood sales through the Internet may be tied to ourestimates o intermediated sales.2 As described in Martinez et al. (2010), theterm local ood sales does not include any geographic reerence. Only 15percent o the respondents reporting direct sales responded in question 28that they used the Internet to directly market their products. Because therewas no ollow-up question related to the volume o Internet sales, responsesto question 28 remain ambiguous and we cannot estimate the volume oInternet sales among the 15 percent o arms indicating local ood participa-tion. As Internet use becomes more ubiquitous in all acets o arm opera-tions, a set o questions will need to be included that identiy Internet sales asa distinct marketing outlet.

2For direct-to-consumer sales, we

avoided this issue by counting only those

producers and the value o their sales

linked to the specic place-based direct

marketing channels in question 23.


27/38

21


MANAGEMENT PRACTICES

21. Next, I have some questions about your marketing practices. Are you currently using CODE

a. direct sales to consumers?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

YES = 11151

b. sales to retail outlets?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1152

c. advisory services?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1153

d. options?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1154

e. futures?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1155

f. on-farm storage?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1156

g. contract shipping (hiring the hauling of your products)?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1157

h. collaborative marketing or networking to sell commodities?. . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1 1158

i. farmer owned co-ops?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1159

ENUMERATOR NOTE: [If 21(a) = 1 or 21(b) = 1 continue, otherwise go to Item 29]

22. In 2008, did you sell products originating from:

CODE

a. crop production other than nursery or floriculture products?. . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1160

(i) did you have these products processed prior to sale?. . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1161

b. nursery and floriculture production? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1162

c. livestock production? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1163

(i) did you have these products processed prior to sale?. . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1164

(ii) if meat products were sold, did you sell by cut of meat?. . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1165

(iii) if meat products were sold, did you sell by fraction of the animal?. . . . . . . . . . . . . . . .YES = 1

1166

d. poultry production? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1167

(i) did you have the birds processed prior to sale?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1168

e. other animal production? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1169

SECTION I continued on next page.

2008 ARMS Questionnaire Questions


28/38

22



23. Did you use any of the following outlets to market these products (Item 22):

CODE

a. roadside stand or on-farm facility (exclude on-farm store)? . . . . . . . . . . . . . . . . . . . . . . . . . . .

YES = 11171

b. on-farm store? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1172

c. farmers market? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1

1173

MILES

(i) If product was sold through a farmers market, what was the distance tothe market where the majority of the product was sold?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1174

CODE

d. Community Supported Agricultural (CSA) buying club? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1 1175

PERCENT

(i) If product was sold through a CSA buying club,what percent of the households/families participating in the club reside in thesame county as your farm?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1176

e. regional distributor? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1 1177

f. State branding program? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .YES = 1 1178

g. direct sales to local grocery stores, restaurants, or other retailers?. . . . . . . . . . . . . . . . . . . . .YES = 1 1179

PERCENT

24. If product was sold through any of the marketing outlets in Item 23, what percentshare of total farm sales did these outlets represent in 2008?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1180

CODE

25. If you sold livestock, poultry, or other animal products (slaughtered animals, milk,cheese, etc.) through any of the above marketing outlets, were facilities availablelocally (within 50 miles) for processing?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . YES = 1

1181

26. Did you collaborate with other farmers to market these products?. . . . . . . . . . . . . . . . . . . . . . . . YES = 1

1182

27. Was your product sold as a farm or regional brand?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . YES = 1

1183

28. Did you use the Internet or mail order to market any of these products?. . . . . . . . . . . . . . . . . . . YES = 1

1184


29/38

23


Appendix 2Modeling Direct-to-ConsumerSales

Prior research has raised many questions about the production and consump-tion o local oods, but relatively ew answers exist (Brown, 2002; Lev andGwin, 2010). Are local oods production and consumption in equilibrium?Are armers producing local ood or nonmarket reasons (e.g., enjoyment,

environmental impact, or as an entry into national marketing)? Data ondirect-to-consumer sales drawn rom the 2007 Census o Agriculture doesallow us to address actors correlated with supply and demand or local oodand its spatial location. We based our conceptual model on the Brown et al.(2006) county-level model o direct sales or West Virginia and extended itto encompass all U.S. counties and commuting zones, while accounting ordetected spatial autocorrelation.

We ound that direct sales are driven by immobile spatial actors correlatedwith growing conditions and marketing opportunities. Results o our county-level and commuting zone-level models supported the regional analysiso 2008 ARMS data and also provided additional evidence o the patterns

detected in the ARMS data. Model results enabled us to draw more data-driven conclusions than the ARMS data alone would have yielded due togeography and the number o observations in the Census o Agriculture thatcannot be matched in the ARMS.

Data

Most studies on the determinants o local oods use survey data (e.g., Morganand Alipoe, 2001; Lyson and Guptill, 2004; Brown et al., 2007), and a ewhave used secondary data (Thilmany and Watson, 2004; Brown et al., 2006).We attempted to model local ood sales using 2008 ARMS data, but thesample size was too small to generate a statistically signicant regression,

precluding testing o the hypotheses. Using ARMS would oer a betterunderstanding o intermediated sales. Due to data limitations, however, wecan only model direct-to-consumer sales. Until additional years ARMS dataare available and we can pool surveys, we use data on direct-to-consumersales to test the location hypothesis, to understand the drivers o local oodsupply and demand, and to provide additional support or our conclusions.

The dependent variabledirect-to-consumer sales o agricultural productsor human consumptioncomes rom the 2007 Census o Agriculture.Explanatory variables were drawn rom the 2002 Census o Agriculture, the2000 U.S. decennial Census, and other publicly available county-level datasources. All variable descriptions, sources, and summary statistics are avail-

able in appendix table 2 or the county-level model and appendix table 4 orthe commuting zone model.

Variables or the commuting zone model dier only or the purposes oaggregating county-level data up to the commuting-zone level. We usedTolbert and Sizers (1996) commuting zones (CZ)the most recent avail-ablebecause they represent an alternative areal unit (or zonal objects)that might be large enough to capture most direct sales transactions (i.e.,production and consumption occur within the same areal unit). As a plausible


30/38

24



spatial representation o commuting and commerce, commuting zones are animprovement over labor market areas because all U.S. counties, including themost rural, are included in a CZ. Additionally, the CZ may be a useul obser-vation unit because, according to the 2008 ARMS, many arm householdswith direct sales have one member commuting to an o-arm job84.8percent o small direct sale arms and 59.5 percent o all direct sale oodarms had at least one member commuting to an o-arm job.

The Model

Our maps and descriptive analysis revealed that substantial variation existsin direct sales across the United States. We used regression analysis tobetter understand how location actors are correlated with direct sales, whilecontrolling or theoretically appropriate variables. Since data to estimatestructural supply and demand equations were not available, we estimateda reduced-orm equation that included supply- and demand-side variables.We used dependent and explanatory variables ound in Brown et al. (2006),changing variables to expand the model rom West Virginia to the UnitedStates. Due to limitations o available data, we cannot include in our modelsall the actors that theory suggests would aect direct sales. The models

assume market equilibrium existsthat supply equals demand.1 We usedcounty-level data as a proxy or local markets and assumed that mostconsumers shop or groceries locally and producers preer to sell locallydue to increasing travel costs or both. We expected, however, that someproducers and consumers traveled outside the county or commuting zone topurchase ood directly rom armers. To account or this county-to-countyor CZ-to-CZ spillover, we utilized a spatial econometric model to controlor any spatial autocorrelation detected in the reduced-orm linear (OrdinaryLeast Squares or OLS) regression.

Hypotheses and Variables

The supply o locally produced ood is determined by producers, but weknow little about producers motivation (Martinez et al., 2010). Fromexisting literature and ARMS data, we know that local ood producers aremost likely to produce ruits and vegetables and be located in or around anurban area (Gale, 1997; Martinez et al., 2010). We hypothesized that:

1. Supply will be higher where there are adequate inputs or production butalso ample marketing outlets, ceteris paribus.

We tested this hypothesis with production variables included in the reduced-orm equations.

Percentage o cropland in the county (PctCropLand) aects supply, particu-larly in urbanized counties where the cost o armland may be high due todevelopment pressure. Percentage o arms with direct sales (DSFarms),or having neighbors who participate in direct sales, is tied to direct saleslevels via neighboring local ood arms. The share o agricultural sales romruits and vegetables (FruitVeg) suggests how amenable the area is to ruitand vegetable production, which is important because ruits and vegetablesrepresent most directly sold ood. The number o armers markets (FmrMkt)represents marketing outlets or producers. We expected all these production-oriented test variables to have a positive relationship with direct sales.

1Unlike many other markets, quantitysupplied and quantity demanded are un-

even at dierent points in the calendar

year and in dierent locales. In North

America, supply o local ood tends to

be high mid-summer to autumn, while

supply is almost zero at other times o

the year, whereas demand tends to be

more constant. By examining annual

direct sales levels, we skirt seasonality

issues. Another unique aspect to local

oods is the local nature o supply.

Using county-level variables enables

us to equate local supply with local

demand, an improvement over State ormulti-State units o observation. The

spatial econometric model specica-

tion also controls or spatial spillovers

between neighboring counties.


31/38

25


We controlled or environmental actors that aect local ood supply.Temperature (JulyTemp) aects the length o the growing season and thecrops that can be grown locally. Topography (Topog) can also aect thecrops being grown locally and the easible scale o agriculture ound in theareal unit. For example, modern large-scale agriculture is easiest where morerelatively fat land is available or cultivation. Thus, we might expect localood production to occur in more marginalized lands that are not as attractiveor large-scale agriculture. Finally, two variablesaccess to interstate high-

ways (Hwy) and high-speed Internet (Internet)were used to controlor built inrastructure, which may aect the supply and marketability olocal oods.

More research on local ood consumers characteristics is available thanon producer characteristics (Brown, 2002; Brown et al., 2006). A body oliterature posits that local ood consumers do not share a particular demo-graphic, but instead are motivated to purchase local oods or the improvedtaste, quality, variety, environmental benets, and/or to support local armers(Brown, 2002; Martinez et al., 2010).

Spatial analysis suggests urbanization may be associated with high levels odirect sales. We hypothesized that:

2. Local oods sales are higher in metropolitan areas than rural becausethere are more consumers.

Since county-level consumer data or the United States are not available,we used county-level averages o supply-side actors2 to test this urbaniza-tion hypothesis, assuming that the supply o local ood will equal demand.We included dummy variables or metropolitan counties (Metro) andnonmetropolitan counties adjacent to metropolitan counties (AdjMetro) withnonmetro, nonadjacent counties as the omitted condition. We also includeddummies or regions because exploratory spatial data analysis ound regional

heterogeneity in direct sales; we included Pacic, Mountain, Midwest, andNortheast while the South served as the omitted condition. We includedpopulation density (PopLand) to control or cross-county population hetero-geneity without distortion. Despite a debate in the existing literature onwhether wealth and income aects local ood consumption, we cannot testwhether demographics or motive aect an individual's desire to purchaselocal oods using county-level means. Rather, we controlled or wealth andincome to the extent possible by using county-level average wage and salaryincome (AveWS) and median home value (MedHomeValue). Due to thenature o commuting zone data (i.e., its aggregation rom county-level data),we instead used per capita income and the maximum o the median homevalue or commuting zones.

Empirical Methods

We rst estimated the models with OLS regression. As is common withthese models, heteroskedasticity was apparent in the initial results (notshown, but available upon request). Ater applying the White-Huber correc-tion and re-estimating the standard errors, we obtained results that remainedconsistent despite the correction. Multicollinearity was not problematic inthe model; the variance infation actors or the explanatory variables were

2Regressors are represented by average

county-wide values so that each county

proxies or the average characteristic

o the pool o local producers and local

consumers.


32/38

26



less than ve. Endogeneity commonly exists in simple, reduced-orm esti-mations like ours, so we used lagged explanatory variables where possibleto reduce this eect. Nevertheless, this endogeneity persists and can causebiased, but consistent, parameter estimates, possibly aecting the p-values oour hypothesis tests. Endogeneity should not alter coecient signs, and weinterpreted our results with due caution.

Given the spatial patterns exhibited as we mapped U.S. direct sales, we tested

the OLS regression or the presence o spatial processes using the LagrangeMultiplier (LM) test. Controlling or spatial processes can reduce statisticalproblems, such as unstable parameters and unreliable signicance tests.Spatial error processes, or nuisance errors, can occur i spatially correlatedvariables are omitted or the value o adjacent observations move togetherdue to common or correlated unobservable variables. Spatial lag processesoccurred due to some systematic interaction among neighbor areal units. Wetested or the presence o spatial processes in both the county and CZ model.

Conducting the LM test requires that an appropriate spatial weights matrix beselected, and there is very little ormal guidance when choosing the optimalspatial weights matrix. We used a rst-order queen contiguity weights matrix

due to the nature o spatial dependence and its suitability or use with irreg-ular polygons. The weights matrix is row-standardized to acilitate interpreta-tion and ease computational expense.

For the county model, we ound both the lag and error LM tests were signi-cant, so we conducted robust LM tests, according to Anselin (1988). Therobust LM tests showed preerence or the spatial autoregressive (SAR)model, suggesting that some sort o systematic county-to-county spatialprocesses exist. When spatial lag processes are unaccounted or in a model,the coecients can be biased and inconsistent, leading to the wrong sign oncoecients and invalid hypothesis testing. Spatial dependence o this typeis most requently incorporated into models using the spatial autoregressive(SAR) lag model:

(1) Y=WY + Y + ,

where ~ i.i.d., W is an n x n matrix dening spatial unit interaction, andthe spatial lag process is accounted or using the spatially-weighted depen-dent variable, WY(Anselin, 1988

USDA: Direct & Intermediated Marketing of Local Foods US

Documents

Transcript of USDA: Direct & Intermediated Marketing of Local Foods US