Beef cattle farmers’ marketing preferences for selling local beef

2.1 Farm characteristics

Kilmer et al. (1994), K. L. Jensen et al. (2010), and Pace and Robinson (2012) found that larger farmers were less likely to be members of cooperatives. Meanwhile, Wachenheim et al. (2001), Puaha (2003), and K. Jensen et al. (2011) found that larger farms were more likely to belong to cooperatives. Gillespie et al. (2016) also found that larger farms were more likely to use multiple marketing channels. While previous research has found mixed results regarding how operation size is related to cooperative membership, we hypothesize larger farms with more animal units will be more likely to prefer to sell finished cattle to a third party for slaughter and be less likely to want to belong to a processing and marketing cooperative (Table 1). We hypothesize this because larger farms would be more able to put together a truckload of cattle to ship to a slaughter facility than smaller operations.

Tennessee is divided into three grand divisions: East, Middle, and West. The West is characterized by row cropping and larger farming operations. The Middle, where the majority of cow–calf and backgrounding operations are located, is characterized by low rolling hills and more moderately sized farming operations. The East is a more mountainous region with smaller farmland parcels. Middle Tennessee has the largest number of beef cattle operations and the largest share of beef cow inventory (58.7%), with the East having the second largest share (30.5%) and the West the smallest share (10.8%; USDA National Agricultural Statistical Service, 2018).2 Since Middle Tennessee has the most beef cattle, they also likely have the most beef cattle farmers and most potential for forming cooperatives compared to the other regions. Therefore, we hypothesize farmers in East and West Tennessee, compared to Middle Tennessee, will be more likely to prefer to sell finished cattle to a third party while farmers in Middle Tennessee will be more likely to prefer to belong to a processing or beef marketing cooperative (Table 1). West, Middle, and East Tennessee were also included in the regression to control for unobserved heterogeneity associated with these regions.

At the time of this study, there were 13 federally inspected livestock slaughter facilities in Tennessee where producers could have their cattle processed and then sell their beef at a retail level to restaurants or directly to consumers (Pepper, Leffew, and Holland,2016). Of these processing facilities, five were located in West Tennessee, four in Middle Tennessee and four in East Tennessee. It is hypothesized that farmers who are located near a federally inspected slaughter facility will be more likely to prefer selling finished cattle to third parties for slaughter or joining a marketing cooperative that markets beef to a third party since they already have access to a processing facility. Meanwhile, we predict farmers who are not located near a federally inspected slaughter facility will be more likely to prefer a cooperatively owned processing facility (Table 1).

The interaction of location and proximity to a federally inspected processing facility is also hypothesized to impact farmers’ preferences for marketing arrangement. In West, Middle, and East Tennessee, 61%, 40%, and 45%, respectively of cattle farmers are located in counties in close proximity to a federally inspected slaughter facility. We hypothesize the farmers located by a federally inspected slaughter facility in East and West Tennessee will be more likely to prefer to sell finished cattle to a third party for slaughter or belong to a beef marketing cooperative since they already have the most access to a processing facility (Table 1).

2.2 Farmer demographics

Studies have found that younger hog farmers (Wachenheim et al., 2001) and wheat farmers (Puaha, 2003) were more likely to belong to a cooperative. Also, Gillespie et al. (2004) found that younger producers were more likely to use private treaties and retained ownership for marketing their cattle than their older counterparts. Therefore, we hypothesize that the older producers are, the less likely they would join the cooperatively owned processing facility and the marketing cooperative, and the more likely they would prefer to market their cattle to a third party (Table 1).

Previous research found that more educated producers were more likely to be cooperative members (K. L. Jensen et al., 2010; Puaha, 2003; Wachenheim et al., 2001). Thus, it is hypothesized that producers with more education will prefer the marketing and processing cooperative compared to selling finished cattle to a third party for slaughter (Table 1). If cattle farmers already sell freezer beef and cattle directly to a packer, it is hypothesized they will prefer to sell finished cattle to a third party for slaughter since they already have a system in place for slaughtering their cattle and selling beef. Thus, the need to belong to a processing or marketing cooperative does not exist (Table 1). No previous research was identified that examined how farm structure impacted cooperative participation. We hypothesize sole proprietor farming operations will be more likely to prefer the processing or marketing cooperative since a sole proprietor may be a smaller operation size than other entity types (e.g., corporation) and prefer to join a cooperative (Table 1).

Puaha (2003) found that wheat cooperative members had a higher share of their income from wheat farming than nonmembers. K. L. Jensen et al. (2010) found that farmers with moderate farm income and high debt were less likely to have an interest in a poultry litter cooperative. K. Jensen et al. (2011) found that farmers with higher off-farm income and moderate debt were more likely to have an interest in a switchgrass cooperative. Given the results indicate that higher incomes are associated with cooperative membership, we hypothesize that farmers with greater income will prefer the marketing or processing cooperative while farmers with lower incomes will prefer to sell their beef to a third party for slaughter. Gillespie et al. (2006) found that producers who farm as a hobby might not want to participate in a beef strategic alliance because they do not wish to devote the time and effort to change management practices. Kilmer et al. (1994) found that as the percent of income from the dairy increases, the more likely the dairy farmer will be a cooperative member. Therefore, we predict that farmers who have over half of their income from beef operations will be more likely to prefer marketing and processing cooperatives while farmers who derive less than half of their incomes from the beef will prefer selling their finished cattle to a third party for slaughter (Table 1). Similarly, we hypothesize that full-time beef producers with higher household incomes will be more likely to prefer marketing and processing cooperatives (Table 1).

2.3 Farmer attitudes

K. Jensen et al. (2011) found that producers who were willing to take more risks had increased preferences for a switchgrass cooperative. Additionally, of the three marketing alternatives, we presented to participants, cooperatively owning a processing facility is the most capital intensive and, thus, riskiest. Therefore, it is hypothesized that farmers who are more willing to take risks will also have increased preferences for the processing cooperatives compared to the other marketing structures (Table 1). As farmers consider using cash forward contracting as an increased barrier, it is hypothesized that farmers will prefer to sell cattle to a third party for slaughter instead of joining a processing or marketing cooperative (Table 1) which is largely due to beef cattle farmers’ desire for marketing flexibility and independent nature. As farmers consider having to accept a price negotiated by a cooperative or marketing alliance as more of a barrier, we predict farmers will prefer not to join the marketing or processing cooperative (Table 1). Finally, as farmers are more likely to agree that consumers would pay a premium for TCB, we hypothesize that they will most likely prefer a marketing cooperative to sell their beef (Table 1).

3.1 Data collection and survey

Beef cattle producers who participated in the Tennessee Agricultural Enhancement Program (TAEP) were surveyed to collect the data used in this study. There were 5,500 beef cattle producers participating in this program spread across the state in 2016. The survey was pretested through email of 275 producers in June and July 2017 and those included in the pretest were not invited to participate in the full survey. The full survey, which included revisions from the pretests, was emailed to 4,661 producers participating in TAEP who had usable email addresses in August 2016. The email contained a link to a Qualtrics survey that producers then completed. A follow-up reminder email was sent a week after the initial email and a second reminder email was sent 2–3 weeks after that. To complete the survey, producers had to raise cattle in 2015 and be the primary decision maker. A total of 864 surveys were collected by mid-September, at which time the survey closed, resulting in a survey response rate of 18.5%.

Next, if the farmer was interested in the TCB Program, they were asked how they would prefer to sell their animals in the program with the options for selling being: (a) sell finished cattle by contract or broker to the third party for slaughter, (b) farmer cooperatively owned processing facility, or (c) farmer marketing cooperative that markets beef to a third party. The exact form of this survey question is available in the Appendix. The final section of the survey asked about farm characteristics, farmer demographics, and farmer attitudes. A copy of the survey is available from the authors upon request.

3.2 Economic model

Following McFadden (1974), a farmer i is hypothesized to prefer the jth marketing arrangement if the utility () exceeds the utility of choosing a different type of marketing arrangement for the branded beef ), where J = 1, 2, 3 (1 = sell finished cattle to a third party for slaughter, 2 = farmer cooperatively owned processing facility, 3 = farmer marketing cooperative that markets beef to a third party) and j ≠ k. Hence, farmer i will choose alternative j if:

(1)

The probability of choosing the jth market arrangement ( is (Greene, 2012):

(2)

Assuming a multinomial logit, then the probability of selecting the jth marketing arrangement becomes:

(3)

The vector of explanatory variables for individual i () is hypothesized to influence the preferred marketing arrangement for branded beef and includes farmer demographics, farm characteristics, and farmer attitudes (Table 1). is the vector of estimated coefficients associated with the explanatory variables.

The marginal effects of the variables are calculated from the estimated coefficients. The marginal effect of the probability of choosing a particular marketing arrangement with respect to each explanatory variable, X, is (Greene, 2012):

(4)

The standard errors around the marginal effects were calculated using the Delta method (Greene, 2012). The marginal effects are estimated by taking the average marginal effect over the estimation sample.

4.1 Preferred marketing arrangements and farm demographics

While a total of 555 beef cattle farmers were interested in participating in the TCB Program, a total of 428 farmers responded to all of the questions needed for the analysis of farmers’ preferred marketing arrangement for selling a branded beef product. Among the marketing arrangements (Preferred Marketing Arrangement), 43.0% would prefer to market TCB through a farmer cooperatively owned processing facility, 38.6% prefer a farmer marketing cooperative that markets beef to a third party, and 18.5% prefer to sell finished cattle by contract or broker to a third party for slaughter (Table 1).

On average, the number of beef cattle animal units managed and marketed (AnimalUnits) was 104.62.4 The Tennessee average herd size is 26 head of cattle; however, this number does not include stocker operations which tend to be much larger than cow–calf operations (USDA Census of Agriculture, 2012). This discrepancy on average herd size may also exist since only cattle producers who had participated in TAEP were surveyed and they may be larger producers than average. About 31.1% of the respondents were located in East Tennessee, while 19.6% were located in West Tennessee, and 49.3% were located in Middle Tennessee. This is consistent with where beef cattle are located in the state since 35%, 11%, and 54% of the cattle and calves in the state are located in East, West, and Middle Tennessee, respectively (USDA National Agricultural Statistical Service, 2018). Just under 46% of producers were located in counties having or bordering a county with a federally inspected slaughter facility (FedInspect). About 14% were in a county in East Tennessee with proximity to a federally inspected slaughter facility (FederalInspect × East), while about 11.9% were located in West Tennessee counties with proximity to a federally inspected slaughter facility (FedInspect × West).

The average age of the responding farmers was nearly 52 years which is consistent with the average age of a US farmer being 58 (USDA Census of Agriculture, 2012). Nearly 58% of the sample were college graduates and the average household income (including farm and nonfarm income) was $131,565. About 8.4% of producers already sold freezer beef through private treaty or sold cattle directly to a packer. Nearly 80% of the respondents were sole proprietors and 17.1% derived at least half of their household income from their beef cattle enterprises (FullTimeBeef).

Producers considered themselves fairly risky when finding new market outlets for beef (RiskNewMkt) with an average score of nearly 8 out of 10. Both potential barriers of having to use cash forward contracting (BarrierForwardCont) or having to accept a price negotiated by a cooperative or marketing alliance (BarrierCoopPrice) were considered moderate barriers with values of just over 2 out of 5. The farmers were in agreement with the statement that consumers would pay a premium for TCB (PremiumTCB).

4.2 Multinomial logit model results

To estimate a multinomial logit model, one of the preferred marketing structure options had to be treated as the omitted category of the dependent variable for the analysis. Thus, private contracting was the marketing structure that was omitted from analysis and selected as the base outcome. However, when presenting marginal effects, this marketing option is recovered from analysis and discussed. The estimated multinomial logit model is shown in Table 2. The log likelihood ratio test shows that the model is significant overall. The model correctly predicts just over 53% of the observations. The mean variance inflation factor (VIF) was 1.81 and the max VIF was 3.53; thus, multicollinearity was not a concern with the model.

In terms of farm characteristics, only the interaction variable FedInspect×East was significant in influencing producers’ preference to join a processing cooperative to market their TCB compared to selling finished cattle to a third party for slaughter. Meanwhile, AnimalUnits and FederalInspect×East were both found to influence a producers’ preference to sell TCB through a cooperative that markets beef to a third party compared to selling finished cattle to a third party for slaughter.

With regard to farmer demographics, Age, PrivTreaty/Packer, FullTimeBeef, and FullTimeBeef×HHInc were all found to influence producers’ preferences to join a processing cooperative to market their beef compared to selling finished cattle to a third party for slaughter. Age and PrivTreaty/Packer also impacted producers’ preferences to join a marketing cooperative to sell their TCB beef. Additionally, HHInc also had an impact on producers’ desire to market their beef through a marketing cooperative as opposed to selling finished cattle to a third party for slaughter.

Farmer attitudes only had an impact on farmers’ preference to join a processing cooperative compared to private contracting finished cattle to a third party for slaughter. If farmers had increased risk preferences (RiskNewMkt) and thought consumers would pay a premium for TCB (PremiumTCB) they were more likely to join a processing cooperative compared to private contracting finished cattle to a third party for slaughter.

4.3 Multinomial logit marginal effects

The estimated marginal effects for the multinomial logit model, which were calculated following Equation (4), are shown in Table 3.

4.4 Quantity of TCB farmers would supply across market arrangements

In addition to asking farmers about their preferred marketing structures for selling local beef, farmers were also asked how much beef they would supply through each marketing channel. Farmers selecting to sell finished cattle to a third party for slaughter would supply, on average, 94,887 pounds of beef per farm. Meanwhile, farms with a preferred marketing structure of processing cooperative were willing to supply, on average, 50,109 pounds of beef, and those choosing a marketing cooperative were willing to supply 52,677 pounds of local beef. This is consistent with our model results that larger farms were more likely to prefer to sell finished cattle to a third party for slaughter (Table 3). Hence, although the majority of farmers would prefer to sell their local beef through cooperatives (Table 1), the quantity of beef per farm they would sell is smaller than for those preferring private contracts.5 When the total quantities of beef were calculated for each type of marketing arrangement across all farms, about 37.14% of the total pounds were estimated to originate from those preferring processing cooperatives, 33.82% from those preferring marketing contracts, and 29.03% from those preferring private contracting.

Dalton, Holland, and Hubbs (2007) surveyed federally inspected facilities that can accept cattle for immediate slaughter in Tennessee and found that respondents slaughtered, on average, 886 head per facility and 7,976 head cumulatively across all facilities. Capacity, however, averaged 3,121 head per plant. Pepper et al. (2016) updated the Dalton et al. (2007) survey and found that for all 13 Tennessee federally inspected facilities cited as accepting cattle for immediate slaughter (including respondents and nonrespondents), this would be around 40,573 head capacity across all facilities and about 11,518 head slaughtered per year. The number of pounds that farmers preferring private contracts would supply was 94,877 pounds per farm annually, with 71 farms choosing this alternative or 6,736,977 pounds across the 71 farms. Assuming 1,300 pounds liveweight per animal, this is about 5,182 head supplied through private contracts. Using the same calculation for marketing cooperatives this results in 6,630 head, and 6,037 head for processing cooperatives, respectively. Hence if the total capacity across the 13 facilities is 40,473 head, with 11,518 head slaughtered per year, it appears that there is likely sufficient existing slaughter capacity in Tennessee to accommodate groups of farmers selling via these marketing arrangements. In particular, selling through private contracts and selling through marketing cooperatives could potentially be accommodated through existing facilities.