Investment Decisions in New Generation Cooperatives:

Transcription

1 Investment Decsons n New Generaton Cooperatves: A Case Study of Value Added Products (VAP) Cooperatve n Alva, Oklahoma* Hubertus Puaha Former Research Assstant Department of Agrcultural Economcs Oklahoma State Unversty Telp. (520) ext Fax. (520) e-mal: hpuaha@u.arzona.edu Danel S. Tlley Professor Department of Agrcultural Economcs Oklahoma State Unversty 422 AgHall Stllwater, OK Telp. (405) Fax. (405) e-mal: dtlley@okstate.edu ABSTRACT: Explanng the phenomena of the new generaton cooperatves development s mportant to understand why some producers nvest n the new generaton cooperatve nvestment and some do not. Results from factor analyss and Tobt model suggest that nonmonetary benefts from nvestment are sgnfcant factors that nfluence producer nvestment decsons n the new generaton cooperatve. KEY WORDS: closed cooperatves, Tobt model, nvestment theory, factor analyss. * Selected Paper prepared for presentaton at the Southern Agrcultural Economcs Assocaton Annual Meetng, Moble, Alabama, February 1 5, Research reported n ths paper was supported by USDA, Rural Busness Cooperatve Servce Grant RBS Copyrght 2003 by Hubertus Puaha and Danel S. Tlley. All rghts reserved. Readers may make verbatm copes of ths document for non-commercal purposes by any means, provded that ths copyrght notce appears on all such copes.

2 Investment Decsons n New Generaton Cooperatves: A Case Study of Value Added Products (VAP) Cooperatve n Alva, Oklahoma Introducton Wthn agrcultural markets n the Unted States, new generaton cooperatves are one of the most mportant new nsttutonal nnovatons. In many states, relatvely conservatve agrcultural producers are nvestng n relatvely rsky new generaton cooperatve ventures. The objectve of ths paper s to explan why some producers nvest n the new generaton cooperatve nvestment and some do not. Throughout the Unted States, many tradtonal cooperatves are mergng, formng jont ventures and coaltons, or strugglng to survve whle new generaton cooperatves are ncreasng n sze and number. Tradtonal cooperatves have struggled to acqure equty because cooperatve ownershp per se conveys no beneft. Benefts generally come only on the bass of patronage. Tradtonal cooperatves attempt to buld equty out of the proft stream. Members receve a porton of ther allocated profts n the form of stock. Generally, there s no secondary market for tradtonal cooperatve stock whch s redeemed at face value by the cooperatve at some future date. New generaton cooperatves attempt to solve the equty problems of tradtonal cooperatves by changng the property rghts structure (Cook and Ilopoulos, 2000). New generaton cooperatves have a more clearly defned membershp polcy (closed or well-defned), a secondary market for members resdual clams, patronage and resdual clamant status restrctons, and an enforceable member pre-commtment mechansm. Oklahoma s frst new generaton cooperatve Value Added Products (VAP) recently opened n Alva, Oklahoma. The cooperatve produces frozen dough products and started operaton n To encourage new generaton cooperatves, the Oklahoma legslature passed 1

3 the Oklahoma Agrcultural Producer Credt Act for Oklahoma agrcultural producers who nvest n Oklahoma agrcultural processng or marketng ventures (68 O.S. Secton ). Ths act allows producers/nvestors to clam an Oklahoma ncome tax credt of up to thrty percent of ther nvestment n Oklahoma producer-owned agrcultural processng cooperatves, ventures or marketng assocatons created and desgned to develop and advance the producton, processng, handlng and marketng of agrcultural commodtes grown, made or manufactured n Oklahoma. Several other groups are organzng to form smlar cooperatves n Oklahoma and throughout the Unted States. Investments n many closed cooperatves may have a hgh degree of rsk. The rsks assocated wth VAP Cooperatve are a promnent consderaton because ths nvestment s a start-up enterprse, whch currently only sells ts product to a lmted number of customers, and ts product market (pzza dough) s n a hghly compettve market. There s drect competton from many companes wth far greater resources and experence. In addton, VAP Cooperatve reles on a sngle product lne and has a lmted product dstrbuton system. Greater understandng of the forces nfluencng new generaton cooperatve development could help exstng cooperatves make changes to survve and facltate the creaton of new cooperatves. Determnants of the survval and stablty of agrcultural closed cooperatves are emprcally tested and evaluated. The model we used n ths paper s an extenson of the prevous theory of agrcultural cooperatves by ntegratng nvestment theory, non-monetary benefts, and farness nto a theory of cooperatve development. Both Staatz (1983) and Sexton (1986) have used cooperatve game theory to study agrcultural cooperatves. Sexton argued that most responses to the forces 2

4 nducng change nvolve the formaton of coaltons 1 that frequently requre fnancal nvestments and have the potental to create non-monetary benefts for members. New generaton agrcultural cooperatves are coaltons of agrcultural producers. The theory of coaltons has been developed largely ndependently n the economcs lterature. The essental dfference between ths paper and prevous studes s that t treats the decson to jon a closed cooperatve as an nvestment decson and suggests that non-monetary payoffs and nvestor s percepton of farness may nfluence nvestment decsons. Closed cooperatve nvestments are consdered wthn the context of a portfolo of nvestment choces a producer can make. A member of a closed cooperatve receves specfc rghts (delvery rghts) n return for hs/her nvestment. These rghts are often transferable and may change n value. Payoffs are based on the amount of nvestment and whether the delvery oblgaton has been met. The value of the delvery rght s expected to be drectly related to both the sze of the monetary dstrbutons to the members as well as the perceved non-monetary benefts created for members. I. The Model For notatonal purposes, we need to defne the varables used n our equatons. Let p = p, K, p ) denotes for the vector prces of the assets. x = x, K, x ) represents the assets ( 1 A ( 1 A or portfolo choces. The varable R = R, K, R ) denotes for expected return on the portfolo ( 1 A choces 1,, A and G = G, K, G ) represents the non-monetary benefts from portfolo x. The ( 1 A nvestor s expected return of portfolo x s denoted by W = Rx ; f s a vector of the nvestors percepton of farness for each asset f = f, K, f ), and W o represents ntal level of wealth. ( 1 A 1 Coaltons n agrcultural marketng systems are horzontal and/or vertcal groups of ndvduals or frms wthn the agrcultural marketng system for whom a new set of bndng rules or contracts are formed. 3

5 U ( ) s the von Neuman-Morgenstern utlty functon whch s enhanced wth non-monetary benefts, rsk, and a farness component. The rsks assocated wth cooperatve nvestment as part of producers portfolo are represented by varance of return on nvestment from the portfolo x. The varance of return from portfolo x s represented by φ x Vx where φ < 0 s the rsk-averson parameter, and the 2 nvestor s utlty from portfolo x has mean µ and varance σ. Utlty s a functon of expected return on nvestment, the varance of return from the portfolo, percepton of farness, and nonmonetary benefts assocated wth that portfolo choce. Producers are hypotheszed to maxmze utlty subject to a wealth constrant: maxu( Rx, φ x Vx, Gx, fx) x subject to p x = W o Suppose that we have observed a portfolo choce choose portfolo x f and only f and x 0 x for U ( Rx, φ x ' Vx, Gx, fx ) U ( Rx, φ x Vx, Gx, fx) = 1, K, n, the ratonal nvestor wll for all portfolo x such that p x p x. Ths expresson tells us that gven the expected return R, varance/covarance matrx V, non-monetary return vector G, and farness vector f, nvestors decde to nvest n the cooperatve membershp f the expected utlty from a portfolo contanng a cooperatve nvestment exceeds any other affordable portfolo. 4

6 There are two ways of provng necessary and suffcent condtons for the valdty of the utlty maxmzaton model 2 : Slutsky condtons and revealed preference condtons (Varan, 1983). Revealed preference condtons are used because ths approach s more applcable for emprcal analyss. The Closed Cooperatve Investment Model The nvestor s nterest s what the optmal value of how the optmal utlty changes as x s to acheve maxmum utlty and 2 x changes. Suppose that µ, D, σ, and F are chosen to maxmze nvestor s utlty functon. For each dfferent value of x there wll typcally be a dfferent optmal choce of 2 µ, D, σ, and F. For example, a dfferent amount of delvery rghts purchased wll determne dfferent optmal choce of monetary and non-monetary benefts, rsks, and percepton of farness. Let us denote the maxmum utlty as M ( x ) for dfferent choces of x, and µ = Rx ; D = Gx ; x 'Vx 2 σ = φ ; F = fx and g( x, W ) p x W 0 = 0, 2 ( µ ( x ), D( x ), ( x ), F( x )) M ( x ) maxu σ x subject to g( x, W0 ) = 0 and x 0 by settng the Lagrangan functon 2 ( µ ( x ), D( x ), σ ( x ), F( x )) λg( x, W ) L( x, λ) = U 0 and takng the frst-order condtons wth respect to nvestment functon, ( 0 (1) x = x R, G, φ V, f, p, W ) x and λ then the closed cooperatve 2 The necessary and suffcent condtons for the mean-varance utlty maxmzaton of closed cooperatve portfolo model are descrbed n Puaha and Tlley (2002). 5

7 Hypotheses The hypotheses generated from our model provde the meanngful reasons why producers nvest n closed cooperatve nvestment: H 1: The producers who want to create employment opportuntes and support economc development n ther local communty are more wllng to nvest n a cooperatve as part of ther portfolo f that nvestment provdes those non-monetary benefts. H 2 : The group of rsk-averse producers s more wllng to nvest n a closed cooperatve f they perceve that nvestment to have relatvely low rsk. H 3 : The producers who are concerned about farness are more wllng to nvest n a closed cooperatve f that enterprse provdes treatment that s perceved as far. II. The Survey Method and Factor Analyss The surveys were sent by mal to 712 members of Value Added Products Cooperatve Assocaton, a closed cooperatve at Alva, Oklahoma and a random sample of Oklahoma wheat growers (members removed) who were non-members of VAP Cooperatve. The survey nstruments for the wheat producers were desgned to allow for comparson of the results between the two samples of wheat producers. The questonnare was frst maled on January 28, One week later, a thank you postcard was maled to all respondents. On February 25, 2002 the second malng of the questonnare was sent out to those who dd not respond n the frst malng. Fnally, those who stll dd not respond receved a phone call requestng completon of the questonnare. Some of the respondents who were called requested a thrd malng. Responses from 298 respondents who dd not nvest and 323 respondents who dd nvest n VAP Cooperatve were receved. 6

8 The VAP Cooperatve questonnare starts wth questons about the respondent s farmland locaton, the length of tme they have operated a farm busness, wheat producton, farm acreage, land ownershp, and some wheat marketng questons. A secton focuses on the respondent s famlarty wth VAP Cooperatve and ther method of learnng about VAP Cooperatve. Respondents were asked about ther expected rate of return on ther VAP nvestment compared to other debt or nvestment nterest rates. Respondents ndcate whether they are able to clam the Oklahoma Agrcultural Producer ncome tax credt as a result of ther VAP nvestment or nvestments smlar to VAP. Then, respondents ndcate whether or not they have off-farm employment. Respondents were also asked to agree or dsagree wth several statements about whether perceptons of farness, non-monetary benefts, tax credt, rsk, marketng contract, and transferablty of VAP s share affected ther nvestment decson. The last part of questonnare ncludes some questons on the respondents demographc characterstcs such as gender, age, and educaton level. The survey of wheat producers produces a complex set of raw data for testng the hypotheses of the proposed model. Raw data consst of several sets of scores of N observatons. A correlaton exsts between sets of scores that can be measured by the correlaton matrx produced. The sets of scores that are recorded from producers atttudes toward the statements about VAP Cooperatve nvestment decsons are grouped by ther classfcaton related to the varables n the model as follow: (1) tems that measure farness; (2) tems that measure atttudes toward the marketng contract; (3) tems that measure socal benefts; (4) tems that measure rsk. In order to smplfy a complex set of data, factor analyss was used. The central dea of factor analyss s to reduce the dmensonalty of a data set that conssts of a large number of nterrelated varables, whle retanng as much as possble of the varaton present n the data set. 7

9 Ths s acheved by transformng the raw data to a new set of varables whch are uncorrelated and ordered so that the frst few factors retan most of the varaton present n all of the orgnal varables. The methods of factor analyss used n ths study are prncpal component and maxmum lkelhood factor analyss. Prncpal component analyss has smple algebra and computaton technques based on how the factors account for varance and explan correlatons. The purpose of prncpal components analyss s to be able to estmate the correlaton matrx, and ths can be done by fndng the characterstc equaton of the matrx. Ths requres two sets of values, the characterstc vectors of the matrx or egenvectors and the characterstc roots or egenvalues 3. Maxmum lkelhood factor analyss, as a method of condensaton, s expected to search for factors. The strongest argument for choosng maxmum lkelhood factor analyss les n the fact that t has statstcal tests for the sgnfcance of each factor as t s extracted. The most crtcal element s whether a factor loadng s sgnfcant or not, regardless of what method of condensaton s used. Normally, a factor loadng of 0.3 that ndcates 9 percent of the varance s accounted for by the factor, s taken as a crteron to ndcate that the loadng s remarkable (Klne, 1994). Ths paper regards a factor as a remarkable loadng f the loadng s above 0.3. Comparable data from members and non-members of VAP Cooperatve were merged nto one data set. The prncpal component analyss s performed by the FACTOR procedure n SAS. The output ncludes all the egenvalues and the pattern matrx for egenvalues greater than one. Gven the sets of scores from producers responses toward the statements about VAP 3 The egenvector s a column of weghts each applcable to one of the varables n the matrx. For example, f there are fve varables there would be fve weghts n the frst vector. The egenvalue s the sum of squares of the factor loadngs of each factor and reflects the proporton of varance explaned by each factor. Thus, the larger the egenvalue the more varance s explaned by the factor. 8

10 Cooperatve nvestment decsons, four socal/non-monetary scores and fve rsk scores were avalable for analyss. Then the hypotheses testng usng maxmum lkelhood method s performed to confrm the number of factors that should be retaned. The combnaton of two methods n ths factor analyss provdes better results because the prncpal component analyss was frst used to get a rough dea of the number of factors before dong the maxmum-lkelhood analyss. Usng the factors generated from the factor analyss, then the model s estmated usng a Tobt procedure that s approprate for the censored dependent varable. The censored regresson model n ths study s estmated usng the method of maxmum lkelhood. Ths model has both dscrete and contnuous parts n ts dependent varable (Johnston and DNardo, 1997). Instead of observng the decson to nvest n VAP Cooperatve, the data on the amount of shares producers nvested are observed. Thus, usng the Tobt model the observed dependent varable s gven by (2) I I = I * = 0 for I for I * * > 0 0 for = 1, K, N * where I represents the amount of share unts producers nvested n the VAP Cooperatve for those who joned the VAP Cooperatve, and zero for those who dd not jon. The estmated equaton s: I (3) = α + α DISTANCE 1 + α RISK α RISK α YEAR + α FAMILIAR 3 + α SOCIAL α RATE 10 + α FAIR 5 + α WORK 11 + α CONTRACT 6 + α TAX 12 + ε for = 1, K, N where mles. DISTANCE s the dstance of respondent s farm locaton from VAP Cooperatve n YEAR s the number of years respondent has farmed, respondent s awareness of the VAP Cooperatve, FAMILIAR s the varable for FAIR s the varable representng the 9

11 respondent s percepton about far treatment delvered by VAP Cooperatve, CONTRACT s the varable representng the respondent s percepton about VAP Cooperatve marketng contract. RISK1 and RISK 2 are the frst-two factors retaned from the maxmum lkelhood factor analyss that represent the respondent s percepton about rsk on VAP Cooperatve nvestment, SOCIAL s the frst factor retaned from the maxmum lkelhood factor analyss that represents the respondent s percepton that VAP Cooperatve creates socal/non-monetary benefts to nvestors, respondent, RATE s the expected rate of return from VAP Cooperatve nvestment for WORK s the dummy varable for off-farm employment, TAX s the dummy varable for the Oklahoma Agrcultural producer ncome tax credt, and ε s an ndependent dentcally dstrbuted error term. The VAP Cooperatve Survey III. The Results Producer characterstcs for those who nvested and those who dd not nvest n VAP Cooperatve are shown n Table I. Seventy-nne percent of the respondents that nvested n VAP Cooperatve were male whle 96 percent were male that dd not nvest n VAP Cooperatve. The mean farm acreage for VAP members was acres wth 39 percent of those acres planted to wheat ( acres) and non-vap members havng an average acres wth 36 percent n wheat ( acres). The VAP members produced an average of 18, bushels n 2000 and 16, bushels n 2001 whle non-vap members produced an average of 10, bushels n 2000 and 9, bushels n

12 Table I. General Descrptve Informaton about Respondents n Study Characterstcs VAP Members Non-Members Gender: Male Female Educaton: Average Hgh school College Post Graduate % % years % % % % 3.77 % years % % % Average Age years years Percentage of ncome from wheat % % Averages: Farm acreage Acres of wheat Farmland was rented from others Wheat producton n 2000 Wheat producton n 2001 Number of years farmng: Average More than 5 years More than 10 years acres acres % 18, bushels 16, bushels years % % acres acres % 10, bushels 9, bushels years % % Famlarty wth VAP Cooperatve s measured on a one to fve scale, wth a one beng not famlar through a fve beng hghly famlar. Forty-three percent of producers that nvested n VAP Cooperatve were moderately famlar wth VAP Cooperatve whle about 48 percent of non-vap members were not famlar wth VAP Cooperatve (Table II). Table II. Percentage of Famlarty wth Value Added Products Cooperatve VAP Members Non-Members Level of famlarty (N=321) (N=280) Not famlar 0.62 percent percent Less than moderately famlar 7.17 percent percent Moderately famlar percent percent Greater than moderately famlar percent 5.36 percent Hghly famlar percent 3.93 percent 11

13 The members share ownershp s shown n Table III. Sxty-eght percent of VAP Member owned between 1,000 to 3,000 shares. About nneteen percent owned between 3,001 to 5,000 shares. Producers that owned more than 20,000 shares were around 0.94 percent. Table III. The Percentage of VAP Cooperatve s share ownershp Percentage of Number of Amount of Shares Responses Responses Between 1000 to 3000 shares Between 3001 to 5000 shares Between 5001 to 7000 shares Between 7001 to shares Between to shares Between to shares More than shares Mnmum VAP Cooperatve s share ownershp s 1000 shares. Results related to producers atttude toward VAP nvestment decsons are summarzed n Table IV. Most VAP members ndcated that VAP Cooperatve creates non-monetary or socal benefts. However, more than ffty percent of non-members dd not ndcate that VAP Cooperatve creates non-monetary benefts (tems a, b, f, and m, Table IV). Eghty-two percent of members and only 37 percent of non-members agreed that creatng jobs n Alva s mportant for them. Ffty-four percent of members sad that other people that they knew were nvestng n VAP. Seventy-three percent of nvestors sad that they knew the people organzng VAP Cooperatve, and 62 percent of them agreed that they would attend the VAP annual meetngs. However, ffty-four percent of non-members stated that the other people that they knew were not nvestng. Sxty-one percent of them dd not know the people organzng VAP, and around fftyone percent would not attend the VAP annual meetngs f they were members. When asked about farness ssues such as treatment of VAP to the nvestor, and dstrbuton of patronage refund, more than 50 percent of members beleved the VAP s treatment and ts patronage dstrbuton were far (tems e and n, Table IV). 12

14 Table IV. Members and Non-members Atttude toward Statements about VAP Cooperatve Investment Decsons Statements VAP Members, n % Non-Members, n % Dsagree Uncertan Agree Dsagree Uncertan Agree a. Creatng jobs n Alva s mportant for me b. Other people I know sad they were nvestng n VAP c. The busness prospectus for VAP appeared logcal d. I could take advantage of the 30% Oklahoma Agrcultural Producer ncome tax credt e. Producers/nvestors n VAP wll be treated farly f. The people organzng VAP were known to me g. Shares n VAP can be bought and sold h. The probablty of patronage refunds would be hgh VAP s a low-rsk nvestment compared to nvestment n farmland j. My other nvestments are low rsk k. The probablty of VAP success was greater than 90% l. Producers need to form cooperatves to ncrease ther ncome m. As an nvestor, I plan to attend the VAP annual meetngs n. The planned patronage dstrbuton from VAP s far o. Marketng/producton contracts are good for agrculture p. Agrc. Marketng coop are better f they have a marketng contract q. Only agrcultural producers are allowed to partcpate n the VAP Coop r. Meetng wheat delvery requrements to VAP s relatvely easy s. Shares n VAP wll apprecate n value Strongly dsagree and dsagree are combned. Agree and strongly agree are combned. Both members and non-members dd not have a problem wth a marketng contracts (tems o and p, Table IV). The rsks assocated wth VAP nvestment showed very nterestng results. Thrty-seven percent of nvestors consdered that VAP Cooperatve was a rsky 13

15 nvestment compared to an nvestment n farmland. Forty-one percent of members and over forty-eght percent of non-members thought that ther other nvestments were hgh rsk. A majorty of non-members were not sure about the rsk assocated wth VAP success n the future (tems h,, j, k and s, Table IV). Investors agreement toward the statement about whether or not they are able to take advantage of the 30 percent Oklahoma agrcultural producer ncome tax credt apparently supports the nvestment hypothess, as may be seen n Table IV, tem d. Maxmum-Lkelhood Factor Analyss The egenvalues ndcate that one factor provdes an adequate summary of the data. One component, wth egenvalue , accounts for 57 percent of the total varance and two components explanng 75 percent of the varance, as may be seen n Table V. Table V. The Egenvalues of the Correlaton Matrx for Socal/Non-monetary Benefts Factors Egenvalue Dfference Proporton Cumulatve The frst factor s a measure of the overall socal or non-monetary benefts factor snce the frst egenvector shows approxmately equal loadngs and has large postve loadngs on all varables (Table VI). The correlaton wth the varable MKW s especally hgh ( ). By takng the average of the squared loadngs of the frst factor, t explans 57 percent of the varance n the correlaton matrx. 14

16 Table VI. The Frst Factor Pattern for Socal/Non-monetary Benefts Varables Varables Descrpton Factor1 JOB Creatng jobs n Alva s mportant to me PIV People that I know also nvest n VAP MKW VAP management are known to me MTG I wll attend the VAP annual meetngs Fgure 1 plots the sze of nvestment as a functon of nvestors socal percepton measures. The sze of nvestment appears to be postvely related to percepton about socal/nonmonetary benefts. Hgher socal factor means more producers perceved that VAP Cooperatve provdes socal/non-monetary benefts. Investment (Shares) Socal Factor Fgure 1: Investors Percepton about Socal Benefts Measure by Sze of Investment The analyss related to the rsk assocated wth the VAP Cooperatve nvestment shows that two factors provde an adequate summary of the complex sets of rsk varables. 15

17 Cooperatve Investment Decsons Usng all measures of cooperatve nvestment decsons, the evdence that perceptons about non-monetary/socal benefts, rsk assocated wth nvestment, and farness affect producers nvestment decsons were tested. The statstcal analyss s restrcted to producers responses and percepton scores avalable from the VAP Cooperatve survey, resultng n a data set of 486 observatons. Accordngly, the LIFEREG procedure n SAS was used to estmate the model. The amount of shares of the producers nvestment vares consderably, wth a mnmum value of 1,000 shares and a maxmum value of 100,000 shares. The mean producer nvestment s 3,589 shares wth a standard devaton of 6, Results n Table VII show that among the explanatory varables, the number of shares producers nvested n VAP Cooperatve s postvely related to FAMILIAR and SOCIAL. The famlarty measure coeffcent s postve and sgnfcant at the1 percent level. Producers who are famlar wth VAP Cooperatve are more lkely to nvest and nvest more. The coeffcent of the socal and non-monetary benefts measure s also postve and sgnfcant at the 1 percent level. Clearly, the results suggest that VAP Cooperatve should create the percepton and the belef that the enterprse produces socal benefts to nvestors. RISK2, whch represents overall responses of producers that predomnantly emphaszes on lowfnancal rsk over the expected monetary return (rsk averse) from VAP Cooperatve nvestment, has a negatve coeffcent and s sgnfcant at the 10 percent level. Large potental nvestors, who are rsk averse, perceve that VAP Cooperatve s a rsky nvestment and wll have less wllngness to nvest n VAP Cooperatve. The number of shares of nvestment are found to be negatvely related to the dstance from Alva (DISTANCE) and off-farm employment (WORK). The result suggests that the key to 16

18 success for VAP Cooperatve nvestment wll be determned domnantly by more full-tme local agrcultural producers support. The farther ther farmland from Alva, the less lkely producers wll nvest n VAP Cooperatve. Potental nvestors are also more lkely to be full-tme farmers. The dstance from Alva (DISTANCE) and off-farm employment (WORK) are sgnfcant at the 1 percent level and the 5 percent level, respectvely. Producers experence n farm busness (YEAR) and marketng contracts (CONTRACT) had the predcted sgn but showed no sgnfcant mpact on VAP Cooperatve nvestment decsons. Producers years of farmng s negatvely related to the VAP Cooperatve nvestment. The coeffcent for CONTRACT s postve but not sgnfcant. Table VII. Parameter Estmate of the Cooperatve Investment Decsons Usng Censored Regresson Model Dependent Varables Lower Left Censored Values 190 I Dstrbuton Normal Number of Observatons 486 Log Lkelhood Noncensored Values 296 Independent Varables Parameter Estmate Standard Errors Constant DISTANCE** YEAR FAMILIAR** FAIRNESS CONTRACT RISK RISK2* SOCIAL** RATE WORK* TAX* ** Sgnfcant at the 1 percent level, * sgnfcant at the 5 percent level 17

19 Farness percepton (FAIRNESS), overall percepton about rsk assocated wth VAP Cooperatve nvestment (RISK1), and expected rate of return (RATE) show predcted sgns, but they are not sgnfcant. The agreement wth the statement that nvestors can take advantage of the Oklahoma Agrcultural Producer ncome tax credt (TAX) shows a postve effect on VAP Cooperatve nvestment decsons. The ncome tax credt (TAX) s sgnfcant at the 5 percent level. Obvously, ths result suggests that Oklahoma ncome tax credt had a postve mpact on the VAP nvestment decson and encouraged the development of VAP. From the results of the VAP Cooperatve nvestment decsons, t s apparent that the nvestors bear rsks due to changes n the relatve busness envronments that drectly affect the VAP Cooperatve as a new enterprse. However, the vast majorty of wheat producers n the Woods County area nvested and became core nvestors n the VAP Cooperatve. The emprcal results gve supportng evdence to explan ths phenomenon. Regardless of the rsks assocated wth VAP Cooperatve nvestment, local agrcultural producers n Woods County nvested because they beleve that VAP Cooperatve generates socal benefts for the local communty. Usng censored regresson procedures, the results show that nvestment provdes socal/non-monetary benefts at the 1 percent level. Usng the evdence from producers response toward socal benefts, ths study fnds that a closed cooperatve can be ntated and wll survve f there s sgnfcant support from local producers concerned about socal/nonmonetary benefts. The Tobt results also found that wllngness to nvest n VAP Cooperatve s less lkely f an nvestor has a strong preference for low rsk nvestments. Producers responses clearly stated that the VAP Cooperatve s not a low-rsk nvestment. Rsk-averse nvestors are not as wllng to be nvestors. 18

20 There s not enough evdence to reject the null hypothess about the mpact of farness on producers wllngness to nvest. IV. Conclusons The evdence examned n the prevous secton s, for the most part, consstent wth the hypotheses developed n Secton 1. The comparson of cooperatve nvestment decsons between VAP members and non-members showed that more explct postve perceptons are requred to convnce producers to nvest. Postve perceptons of local producers provded the support the VAP Cooperatve needed to be developed. And even though many local producers nvested, the local producers clearly dd not beleve that VAP Cooperatve was a low-rsk nvestment as compared to nvestment n farmland. A hypothess test confrmed that socal or non-monetary benefts have sgnfcant mpacts on cooperatve nvestment. The results suggest that a new generaton cooperatve needs strong support from local producers as core-nvestors to ntate and mantan cooperatve as an operatonal busness. Producers who are famlar wth VAP Cooperatve were more wllng to nvest n VAP Cooperatve, and producers wth farmland far away from Alva dd not nvest n VAP Cooperatve. Strong preferences for low-rsk nvestment lowered producers wllngness to nvest n VAP Cooperatve. Wth regards to farm-employment status, full-tme farmers showed a greater ntenton to nvest rather than part-tme farmers. 19

21 REFERENCES Cook, Mchael. Major Forces n the Agrbusness Envronment of the 1990s. Internatonal Agrbusness Management Assocaton Inaugural Symposum 1991, Proceedngs: Cook, Mchael L. and C. Ilopoulos. "Ill-Defned Property Rghts n Collectve Acton: The Case of US Agrcultural Cooperatves" n C. Menard (ed.) Insttutons, Contracts and Organzatons, London, UK, Edward Elgar Publshng, Greene, Wllam H. Econometrc Analyss. Thrd Edton. New Jersey: Prentce-Hall, Inc., Jollffe, I.T. Prncpal Component Analyss. New York: Sprnger-Verlag, Inc., Klne, Paul. An Easy Gude to Factor Analyss. London: Routledge, Ladd, George W. A Model of a Barganng Cooperatve. Amercan Journal of Agrcultural Economcs (August, 1974): Puaha, Hubertus and Danel S. Tlley. Coalton Development n the Agrcultural Marketng System. Electronc document. Avalable at: umn.edu/cg-bn/pdf_vew.pl?paperd=4376&ftype=.pdf SAS Insttute Inc. SAS/ETS User s Gude. Verson 6. Frst Edton. Cary, NC: SAS Insttute Inc., Sexton, Rchard J. Imperfect Competton n Agrcultural Markets and the Role of Cooperatves: A Spatal Analyss. Amercan Journal of Agrcultural Economcs 72 (August, 1990): The Formaton of Cooperatves: A Game-Theoretc Approach wth Implcatons for Cooperatve Fnance, Decson Makng, and Stablty. Amercan Journal of Agrcultural Economcs 68 (May, 1986): Staatz, John M. Farmer Cooperatve Theory: Recent Developments. USDA, ACS Research Report No. 24, Washngton, D.C. June 1989, 30pp.. The Cooperatve as a Coalton: A Game-Theoretc Approach. Amercan Journal of Agrcultural Economcs 65 (December, 1983): Varan, Hal R. Nonparametrc Tests of Models of Investor Behavor. Journal of Fnancal and Quanttatve Analyss 18 No.3 (September, 1983): Varan, Hal R. Mcroeconomc Analyss. Thrd Edton. New York: W.W. Norton & Company,