thb4393.tmp A Financial Training Program for USDA/FSA Borrowers: Evolution and Impacts Robert L. Parsons, Gregory D. Hanson, Wesley N. Musser, Roland Freund, and Lehan Power A financial training program designed by Cooperative Extension specialists was provided to over 2,000 USDA/FSA borrowers from the Northeast during the period 1994-1999. Key to the success of the workshops was an in-depth, user-friendly curriculum that evolved over time, eventually replacing satellite-feed instruction with pre-taped videos. Cluster analysis classified nearly 70% of workshop participants as “LOW Fkrance Priority” or “Low Fhance Knowledge.” Farmers in these clusters received a relatively greater educational benefit from the program than those not in these clusters. Impact analysis indicated that perceived annual gain in farm net worth from application of workshop tools ranged from approximately $5,000 to $10,000. The training addressed the needs of prodncers typically isolated from Cooperative Extension because the workshop was the only extension program attended that year by nearly two-thirds of them. More than 2,000 farm operators, partners, and spouses in Delaware, Maryland, New York, and Pennsylvania completed a financial training pro- gram developed in response to a mandate in the 1991 U.S. Farm Bill requiring certain farmers with U.S. Department of Agriculture Farm Service Agency (FSA) operating andlor ownership loans to receive instruction in financial statements, budget- ing, record keeping, and financial management practices (Federal Register 1993; Hanson 1995 (table 1)). In order to make the program more use- ful for FSA borrowers, the curriculum incorporated the FSA farm record system and accompanying financial statements format. 1 Because FSA is a lender of “last resort” to limited-resource produc- ers, Cooperative Extension has historically viewed FSA clientele as critically important to its outreach mission (Hanson 1997). The purpose of this paper is to describe the evo- lution and evaluate the effectiveness of this finan- cial training program. Analysis is based on partici- pant evaluations of the program and information about the participants collected during the work- shops. In addition to tabulations of these data, clus- ter analysis and logit models are used in the analy- sis, This analysis is combined with descriptions of the evolution of the program. Initial Curriculum and Workshops Robert L. Parsons is an assistant professor with the Dep~ment of Com- munity Development and Applied Economics at the University of Ver- mont. Gregory D. Hansen is an associate professor with the Department of Agricultural Economics and Rural Sociology at the Pennsylvania State University, ‘ This program is unique in its adoption of the USDAffSA record and financial statement format. Panly as a result of the dkect linkage to the FSA record system and its success in the Northeast, this finance curricu- lum was selected for a nationwide program of training for more than 1000 FSA farm loan officers, county executive directors, and district directnrs frnm 1997-99. The financial training workshops initially used sat- ellite down-link presentations that would be coor- dinated locally by an on-site extension agent. Par- ticipants followed the satellite presentations and did exercises in their own workbooks. With an educationally diverse audience in mind, the cur- riculum emphasized practical applications and had minimal narrative. This format facilitated in-class Agricultural and Resource Economics Review 29/2 (October 2000) 240-250 Copyright 2000 Northeastern Agricultural and Resource Economics Association Parsons et al. A Financial Training Program 241 Table 1. Northeast Finance and Production Training: Farms, Attendees, 1995-99 Basic Cost cost Workshop Farmers, Farms per Farma Per Spouses, Training Enrolled PA/Non-PA Person Curriculum Partners Year PA/non-PA ($) ($) PA/non-PA FINANCE 1994/95 139 290/290 156 259 1995/96 181/12 290/290 216 239/16 1996/97 173/46 190/1 90 132 251/54 1997/98 25133 90/190 120 25146 Finance Total 1998/99 1995-99 378/37 1,024 90/190 182 491/56 1,437 PRODUCTION (PA) 1996/97 188 75 51 279 1997/98 225 ~ ~ ~ Production Total 1996-98 m 87 59 599 FINANCE and PRODUCTION TOTAL 1995–99 1,437 $155 $109 2,036 Source: Short Course Office, Penn State College of Agricultural Sciences. ‘The basic charge does not include late fees and charges for more than two participants per farm. The $290 tuition fee included $90 for meats. The 1997/98 reduction in tuition to $90 in Pennsylvania reflects lower costs due to economies of size. The 1997/98 Pennsylvania finance training was a make-up program, Since 1997/98, Pennsylvania has rotated finance and production training in alternating years. The production training is presently offered only in Pennsylvania. presentations via satellite of topics such as produc- tion-based accrual income. Major sections of the curriculum were “the balance sheet,” “the income statement,”” the cash flow budget,” “financial ratio analysis,” “farm home budgeting,” “strategic plan- ning,” and “fixing broken finances.” The work- book was prepared for a ninth-grade reading level to facilitate the participation of Old Order Amish producers, who do not attend high school, and other farmers with limited educational achieve- ment. However, sufficient conceptual depth was included to challenge college-educated producers. A benefit of keeping the narrative to a minimum was that borrowers took ownership of the text by highlighting and writing notes pertaining to key finance concepts in the text. In order to promote attendance and minimize participant travel time, concurrent workshops were scheduled in approximately 20 different accessible locations. Workshop duration was five to six hours per day, for six days. Agronomy, livestock, and farm business management agents were trained as workshop site leaders. In the first year of the pro- gram, an extension finance specialist presented the text material via satellite up-link from Penn State. This approach assisted the site instructors, many of whom had limited finance background, as well as promoted uniform teaching, No one single special- ist could have delivered all of these lectures on site. The extension agents on-site coordinated training facility logistics and led text exercises, homework, and quizzes. The workshop began with a pretest, followed by text instruction on financial concepts and statements, numerical exercises, and quizzes after completion of each major topic. Each partici- pant was required to complete a balance sheet, an accrual income statement, and a projected monthly cash flow for the coming year. Grades were “Pass,” “Pass with additional FSA-led training re- quired,” and “Fail,” and were based on attendance, effort on exercises, quizzes, and completion of own-farm homework. A panel of experts partici- pated in two live satellite question-and-answer ses- sions that permitted participants to call or fax ques- tions to Penn State. Evolution of Instruction Methods The official FSA evaluation indicated that 87% of the participants found the topics covered in 1994/ 95 to be helpful to the farm business (table 2). While coverage and suitability of the material were considered excellent by only 30?Z0and 29T0 of the producers, respectively, approximately 80% found that the course level, course length, and amount of outside work were “appropriate.” The percentage of respondents who gave ratings of “poor,” “too easy,” or “too short” ranged from only O to 570. Post-workshop discussions between the site leaders and the extension specialist leading the program revealed dissatisfaction with the rigid schedule of satellite up-links. Satellite instruction required that each site meet at the same time/date and complete workshop exercises on a tight sched- ule. Another problem was that signal reception was interrupted at several sites because of equip- ment failure. Accordingly, time dedicated to down- 242 October 2000 Agricultural and Resource Economics Review Table 2. FSA/USDA Financial Management Training Participants Evaluations, 1994-99 DE, MD, DE, MD, Response to PA MD, NY, PA NY, PA PA Evaluation Evaluation 1994/95 1995/96 1996/97 1998/99 Item Item (%) (%) (%) (%) Number of workshop participants 195 265 211 383 comdeting evaluations 1. 2. 3, 4. 5. 6. 7. 8. 9. — Top;cs c&ered in the class Yes 87 88 80 84 were helpful to me in my Partially 13 12 20 16 business No o 0 0 1 Coverage of the subject matter Excellent 30 55 56 48 was Sufficient 67 45 44 52 Poor 3 0 1 1 Suitability of the instruction Excellent 29 48 43 44 materials was Sufficient 71 51 52 54 Poor o 1 5 3 The level of the course was Too advanced 16 6 6 12 Appropriate 83 93 91 86 Too easy 1 1 3 2 The length of the course was Too long 17 7 22 21 Appropriate 78 89 73 74 Too short 5 4 5 5 The amount of outside work Too much 9 6 14 15 was Appropriate 87 91 84 82 Too little 4 3 2 3 The instructor(s) was Excellent 30 55 76 64 Good 66 44 24 35 Poor 4 1 0 1 Will you continue to take Yes 38 36 30 22 training courses in production Maybe 56 54 57 55 and financial management No 6 10 13 24 topics if not required? Would you recommend this Yes 74 85 92 89 instructor to other individuals? No comment 22 12 7 11 No 4 3 1 1 links was reduced from six four-hour sessions in the first year to two one-hour sessions in the third year. As with other distance education programs at Penn State (Peterson 1999) and in other states (Hiel and Herrington 1997), distance education via satellite up-links had proven to be too cumber- some, rigid, and expensive compared to pre-taped video presentations, and so instruction by satellite was finally discontinued altogether in 1998–99. This evolution was beneficial mostly where the local down-link facility was inconveniently located for participants, too small to accommodate all par- ticipants, and/or where local scheduling conflicts existed. Cost savings generated from elimination of satellite instruction helped to reduce tuition from $290 to $90 in Pennsylvania, and to $190 in surrounding states. On-site extension agents, aided by pre-taped instructional videos, provided more of the instruction themselves. These changes permitt- ed site leaders to exercise more control over the pacing of materials. The site instructors, whose knowledge about the subject had increased during the first two years of the program, felt confident about their ability to assume more instructional re- sponsibility after the satellite program was discon- tinued. Increased instructor experience and the use of videos and on-site instruction probably all con- tributed to enhanced instructor ratings over time (table 2, item 7). Beginning in 1995/96, participants were re- quired to complete a four-year farm plan that in- cluded projected yields, expenses, revenues, pro- jected capital expenditures, and family living expenses. In addition, the own-farm financial state- ments that participants were required to prepare were made more challenging. Text workbooks and farm plan booklets were revised annually, not only to keep them current but also to add improve- Parsons et al. A Financial Training Program 243 Table 3. Evaluation Results from 1995-99 FSA Finance Workshops (PA, NY, MD, DE) Mean Values for Participants by Year Participant Characteristics and Evaftration Items 1995/96 1996/97 1998/99 1 2. 3. 4, 5, 6. 7, 8. 9, 10. 11. 12. 13. 14. 15. 16. Number of participant evaluations 180 70.0 76.7 $185,470 151 66.2 75.9 $171.520 364 73.1 75.7 $170,970 Dairy major farm enterprise (%) Number of cows in dairy herd Annual farm sales (including contract income) ($) Age (yrs) Years managing a farm 42.8b 16.7’ 87.2h $19,310 1.Za,b 42,7’ 15,5 86.8’ $18,200 o.9a 39.6bC 14.1b 73.4b,” $17,740 1,0’ Completed high school (%) Annual avg. farm profit past 3 years ($) Change in view of importance of financial management (Scale 1 to 5) Change in knowledge level of farm financial statements (Scale 1 to 5) 1,4...1.5” lff Change in knowledge level of farm financial plans (Scale 1 to 5) 1,5 1.4 1.5 Budgeting, analysis, and planning tools from workshop will help your farm to survive (Scale 1 to 5) 3.9 4.0 3.9 Satisfaction with financial workshop (State 1 to 5) 4.0 $7490 4.0 $6900 3.9 $7330Financial skOls learned in this workshop will likely increase your farm’s net worth per year ($) Training-related increase in net worth as percent of sales (item 14 divided by item 4) 6.3 72.6 7.0 64.6 7.6 68.6Participants not attending other extension workshops in past year (%) ‘Statistically significant difference between 1995–96 and 1996–97 at p <0.05 level, bStatistically significant difference between 1995–96 and 1998–99 at p <0,05 level. ‘Statistically significant difference between 1996–97 and 1998–99 at p <0.05 level, ments suggested by site leaders and participants. For the same reasons, the instructional video tapes used in the third year were also remade, including being shortened by 30-40%. Pennsylvania and Maryland extension specialists and agents were the instructors on these revised tapes. In 1997/98, Farm Production Management was taught instead of finance. This one-year break in the finance curriculum permitted an in-depth revi- sion of its text and instructional format, to correct mistakes and clarify material. These changes in the instructional format resulted in improved evaltta- tions, the “coverage of subject matter” and “suit- ability of instruction material” receiving an “excel- lent” rating by a respective 55% and 48% of the participants (table 2). Although unfavorable ratings increased over time in three of the evaluation cat- egories (on the length of the course, the amount of work required outside the classroom, and the amount of interest in taking additional courses on the same subjects if not required to do so), the rest of the categories, including those pertaining to in- structors, continued to receive favorable ratings. An additional evaluation instrument was added in 1995/96. The purpose of this instrument was to provide information on participant characteristics, change in knowledge levels, and perceived poten- tial impact of the training on net worth accumula- tion (table 3). Three items (9–1 1) are producer as- sessments of their beginning and ending knowl- edge levels of financial topics, and four items ( 12– 15) are self-assessments of workshop satisfaction and impacts. It is important to note that changes in knowledge levels and impacts of knowledge are 244 October 2000 Agricultural and Resource Economics Review Table 4. Evaluation Results by Education Level, 1998-99 FSA Workshops (PA, MD, DE) Mean Values for Participants by Education Level Participant Characteristics and Evaluation Items Primary High School College 1. 2. 3. 4. 5, 6. 7. 8. 9. 10. 11, 12. 13. 14. 15. Percent of participants Dairy major farm enterprise (%) Number of cows in dairy herd Annual farm sales (including contract income) ($) Age (yrs) Years managing a farm Annual avg. farm profit past 3 years ($) Change in view of importance of financial management (Scafe 1 to 5) Change in knowledge level of farm financial statements (Scale 1 to 5) Change in knowledge level of farm financial plans (Scale 1 to 5) Budgeting, anatysis, and planning tools from workshop will help your farm to survive (Scale 1 to 5) Satisfaction with financial workshop (Scale 1 to 5) Financial skills learned in this workshop will likely increase your farm’s net worth per year ($) Training-related increase in net worth as percent of sales (item 14 divided by item 4) Participants not attending other extension workshops in past year (Yc) 24.4 92.0a’b 51.5’J’ $127,920’ 32.7a’h g,ja.b $14,430’ 1,2 1.5 1.5 3.9 3f+b $6520 6.3 73.3 58.0 70,7”’” 89,3’ $185,800’ 42.2’ 16.5’ $20,350” 1.0 1.6 1.4 4.0 3,9a $7900 7.7 69.1 17,6 54,8b’c 75,6b $184,200 40.9b 14.1b $12,010C 0.9 1.7 1.7 4.0 4. lb $6720 6.4 61,7 ‘Statistically significant difference between primary and high school education at p <0.05 level. bStatistically significant difference between primary and college education at p <0.05 level. “Statistically significant difference between high school and college education at p <0.05 level, difficult for both resident and extension educators to assess. However, the consistency of the self- assessment scores over the years supports the view that workshop participants experienced little diffi- culty answering the impact questions. Data in table 3 indicate that the typical partici- pant had managed a farm for about 15 years, was about 40 years old, and had annual farm sales of approximately $ 170,000–$ 185,000. The view that financial management was important and the knowledge levels of farm financial statements and farm financial plans all increased substantially af- ter taking the course (table 3, items 9–11 ). The rating of 3,9-4.0 (on a scale of 1 to 5) indicated that the participants believed that the tools learned at the workshop would help their farms to survive. Participants estimated that implementing the work- shop farrn/household analysis and planning tools could increase farm net worth by an average of about $7,000 in a typical year. As shown by a rating of 3,94.0 (on a scale of 1 to 5), the partici- pants expressed a high degree of overall satisfac- tion with the workshop, The information provided on this evaluation also suggests that the training addressed the needs of producers typically isolated from Cooperative Extension—the workshop was the only extension program attended that year by nearly two-thirds of them. Workshop Evaluations by Education Level and Farm Size Using data from the evaluation instrument, evalu- ations were tabulated based on educational level (table 4) and farm size of the participants. Partici- Parsons et al. A Financial Training Program 245 Table 5, Evaluation Results from 1998-99 FSA Finance Workshops by Gross Sales Mean values for participants by farm sales Sales less than Sales $100,000 Sales greater Participant Characteristics and Evaluation Item $100,000 to $199,999 than $200,000 1. 2, 3. 4. 5. 6. 7. 8, 9. 10. 11. 12. 13, 14. Percent of participants Dairy major farm enterprise (%) Number of cows in dairy herd Annual farm sales (including contract income) ($) Age (yrs) Years managing a farm Completed high school (%) Annual avg. farm profit past 3 years ($) Change in view of importance of financial management (Scale 1 to 5) Change in knowledge level of farm financial statements (Scale 1 to 5) Change in knowledge level of farm financial plans (Scale 1 to 5) Budgeting, analysis, and planning tools from workshop will help your farm to survive (Scale 1 to 5) Satisfaction with financial workshop (Scale 1 to 5) Financial skills learned in this workshop 36.3 60,5a’b 55.9b $55,750”b 37.7b 11.9b 79,5 $12,610a’b 1.2b 1.7 1.6b 3.9 3.9 $5300”b 37.3 83.8’ 59.8C $138,500’” 39,4C 13,7C 64.9 $17,530’” 1.0 1.5 1.4 3,9 3.8 $8050” 26.4 81.9b 126.0b’c $379,1 10”C 43.1b’c 18.1b” 86.6 $24,610b’c 0,8b 1.5 1.3b 4,0 4.0 $lo,150b will likely increase your farm’s net worth per year ($) 15. Training-related increase in net worth as 12.2’J’ 6,0W 36b,C percent of sales (item 14 divided by item 4) 16, Participants not attending other extension 65.4 69.4 70.0 workshops in past year (%) ‘Statistically significant difference between low sales and medium sales at p <0.05 level. bStatistically significant difference between low sales and high sales at p <0.05 level. ‘Statistically significant difference between medium sales and high sales at p <0.05 level pants in 1998-99 who had completed at least high school reported approximately $185,000 gross rev- enue, However, those who had completed high school but not college were more specialized in dairy, had more cows per herd, and reported about $8,000 more in profit than those who had gone to college, The greatest change in views of the im- portance of financial management was shown by the lowest education group, which included the Amish farmers. The change in knowledge vari- ables and satisfaction with the workshop tended to increase with education level, As expected, col- lege-educated participants attended more extension meetings, Overall, the evaluations indicate that the training had similar impacts on knowledge levels for participants at all education levels, even though the workshop experience was more satisfactory for better-prepared college-educated participants. Evaluations were also tabulated by the amount of gross sales reported by the participants into three groups—those reporting sales greater than $200,000, $100,000-$199,999, and less than $100,000 (table 5). The one with the largest gross sales reported less knowledge gain in terms of statements and plans than did the group with the least sales. The higher-sales group did find the workshop slightly more satisfactory and slightly more beneficial in terms of helping their business survive than did the other group, even though the difference was not statistically significant. The 246 October 2000 Agricultural and Resource Economics Review Table 6. Evaluation Characteristics from 1998-99 FSA Finance Workshops by Cluster Cluster Low Finance Low Finance High Finance Priority Knowledge Knowledge Cluster 1 Cluster 2 Cluster 3 Participant Characteristics and Evaluation Items (n = 135) (n = 113) (n = 114) 1. 2. 3. 4. 5. 6. 7. 8. 9. Io. Il. 12. 13. 14, 15. 16. 17, — Percent of participants 37,3 Dairy major farm enterprise (%) 73.3 Number of cows in dairy herd 74.4a’h Annual farm sales ($) 163,047’” Age (yrs) 40.3 Years managing a farm (yrs) 14.0 Completed high school (%) 74.2 Annual avg. farm profit past 3 years ($) 15,682b Change in view of importance of , ,Sa.h financial management (Scale 1 to 5) Change in knowledge level of farm financial I .7h statements (Scale 1 to 5) Change in knowledge level of farm financial 1.6b plans (Scale 1 to 5) Budgeting, analysis, and planning tools from 4.1” workshop will help your farm survive (Scale 1 to 5) Satisfaction with financial workshop (Scale 1 to 5) 3,9 Financial skills learned in this workshop will likely 7233’ increase your farm’s net worth per year ($) Training-related increase in net worth as percent of 9,0d.h sales (items 14 divided by 4) Participants not attending any other 67.4 extension workshops in past year (%) Percent of Amish participants 17.8 ‘Statistically significant difference between cluster 1 and cluster 2 at p <0.05 level. !Statistically significant difference between cluster 1 and cluster 3 at p <0.05 level. ‘Statistically significant difference between cluster 2 and cluster 3 at p <0.05 level. group with the highest gross sales also gave the potential impact of the workshop on annual growth in farm net worth the highest dollar value. How- ever, the group with the lowest amount of gross sales gave a higher rating than the other groups for the potential of the workshop to increase net worth as a percent of sales. The group with highest gross sales entered the workshop better prepared; their pre-workshop scores for items 9–11 were each about 0.4 larger than for the groups with the lowest in gross sales. The post-workshop scores for these items were only about 0.2 larger for the highest gross sales group than the group with the smallest sales. Thus, a general conclusion is that the train- ing succeeded for all education and farm sales levels. 31.2 69.0 65.5’” 128,626’” 38,1 12.7 75.0 15,006’ O&,. 2.0’ 1.9C 3.8” 3,9 6778’ 6.8’ 62.8 20,4 31.5 76.3 86@.c 225,22 I b“ 40,2 15.8 78,0 22,285h’C (33hs , ,Ob,c ~,sb,c 3.9 3.Y 8077bC 6.5b 58.8 16,7 Cluster Analysis of Workshop Participants Tabulations by single variables were helpful in evaluating the success of the workshops, However, analysis for groups defined by several variables further refined the evaluation. Cluster analysis was utilized to delineate groups in a multivariate frame- work for further analysis. Variables used in the cluster analysis were pre-workshop beliefs partici- pants had on several topics: (1) their view of the importance of financial management, (2) their knowledge about farm financial statements, and (3) their knowledge about farm financial plans. Changes in these variables (items 9-11, table 3) were tabulated in tables 5 and 6). The approach of cluster analysis is based on the view that partici- Parsons et al. pants with similar perspectives and knowledge lev- els could be characterized by a similar set of char- acteristics (Bernhardt et al. 1996). It is assumed that participant knowledge can be analyzed in terms of clusters that broadly share similar charac- teristics as C1=fl(BJ, k=l,2, . . ..K C2=f2(BJ,l =l,2,. ... L C~=f~(BZ), z=l,2, . . ..Z. where Ci(i = 1,2, . . . , M) represents the ith clus- ter and Bj Q = k,l, . . . z) is a set of characteristics associated with the ith cluster. These cluster pro- files are mutually exclusive. The FASTCLUS pro- cedure in SAS (SAS/STAT Users Guide 1989) was used to determine the number of clusters and to group the participants. The three clusters identified in the analysis can be described as “Low Finance Priority,” “High Fi- nance Knowledge,” and “Low Finance Knowl- edge” (table 6).’ The 135 participants identified with a Low Finance Priority had an average score of 2.59 (scale of 1–5) on their initial view of the importance of financial management. However, this group had the largest change in their view of financial management. The Low Finance Priority group also had a strong belief that the financial tools acquired in the workshop would help their farms to survive, with an average score of 4,1 on a scale of 1 to 5. Given their initial low priority for finance, it is not surprising that the change in fi- nance knowledge was substantial for the Low Fi- nance Priority group< The 113 members of the Low Finance Knowl- edge cluster had the largest increase in knowledge of financial statements and planning and estimated that use of workshop concepts would raise annual net worth by 6.8 percent (item 15, table 6). The sales level of this group indicates the presence pri- marily of small farmers. The High Finance Knowl- edge cluster scored the lowest on change in view that use of workshop tools would contribute to farm survival, and they also had the smallest in- crease in knowledge of financial statements and planning. Most importantly, the post-workshop view of the importance of financial management converged between 4.4 and 4,8 for the three clus- ters, and knowledge of financial statements and plans ranged from 3.8 to 4.3, suggesting that the workshop tended to make the ending finance knowledge and finance perspective similar for the three clusters. The cluster analysis isolated the par- ticipants with low finance knowledge and a nega- tive belief in the importance of finance. These two groups would be expected to gain less from the workshop than individuals with more knowledge A Financial Training Program 247 and/or more positive beliefs. The fact that their post-workshop knowledge and beliefs had become nearly as high as the group with higher knowledge initially indicated that the curriculum allowed these potentially problem participants to fully par- ticipate and become finance-literate. Logit models (Madalla 1983) were estimated for further comparison of each cluster to the other two clusters. These models considered characteristics of each cluster in a multivariate framework rather than in the univariate tabulations discussed above. Each model has the same set of explanatory vari- ables, which are defined as being 1 when the group of participants (cluster i) has the characteristics, and O when it does not: 10g [pi/(l – ‘i)] = PO+ ~~X~ + ~~X~ + . . . + (31bXlCj+ ei, where Pi = probability that the respondent belonged to the group (cluster) xl = XJ= X3= X4= X5= X6= X7= Xs = X9= cha~ge in the view of the importance of financial management (scale of 1 to 5) change in the level of knowledge of farm financial statements (scale of 1 to 5) change in the level of knowledge of farm financial plans (scale of 1 to 5) 6-10 years of farm management experience more than 10 years of farm management experience moderately satisfied with workshop (score of 4 on scale of 1 to 5) highly satisfied with workshop experience (score of 5 on scale of 1 to 5) farm sales greater than $100,000 off-farm income of $1–$7500 Xlo = off-farm income greater than $7500 Xl ~ = workshop skills will increase net worth $1-$5,000 X12= workshop skills will increase net worth more than $5000 X13 = typical profit $0-$10,000 Xlq = typical profit greater than $10,000 X15 = did not attend any other extension workshops in past year X16 = Amish farmer. The regression coefficients are in the appendix. Given that the explanato~ variables are binary, odds-ratios were computed instead of marginal probabilities (Hosmer and Lemeshow 1989), These odds are used to analyze differences in character- istics among the clusters. The odds of an outcome being present when a predictor variable (X) is equal to one is defined as T(l)/[(1 – T(1)], The 248 October 2000 Agricultural and Resource Economics Review Table 7. Logistic Regression Odds Ratio from 1998-99 FSA Finance Workshops by Cluster Low F]nance Low Finance High Fhrance Priority Knowledge Knowledge Regression Variable Cluster 1 (n = 135) Cluster 2 (n = 113) Cluster 3 (n = 114) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11, 12. 13. 14. 15. 16. 17. — Intercept Change in the view of the importance of financial management Change in the knowledge level of farm financial statements Change in the knowledge level of farm financial plans Farm management experience: 6–10 yrs Farm management experience: more than 10 yrs Moderately satisfied with workshop Highly satisfied with workshop Farm sales greater than $100,000 Off-farm income: $1-$7500 Off-farm income: more than $7500 Workshop skills will increase farm net worth between $ 1–$5000 Workshop skiUs will increase farm net worth by more than $5000 Typical profit: $0-$10,000 Typical profiti greater than $10,000 Participants not attending any other extension workshops in past year Amish producer 0.168 0.073 7.363 13.204*** 0.205*** 0.171*** 0.527** 4.350*** 0.289*** 0.741 2.722*** 0.329*** 1.166 0.706 1.824 1.456 0.775 1.001 0.453** 0.666 4.264*** 0.368** 0.708 5.086*** 0.955 0,630 2.214 0.711 1.314 0.884 1.017 1.681 0,497* 1.525 0.672 1,647 1.153 1.082 0.808 1.222 0.661 1,277 1.087 0.560” 1.601 1,115 0.932 0.858 0.618 2.119* 1.174 *Logistic regression parameter estimate statistically significant at p <0.10 level. **Logistic regression parameter estimate statistically Significant at p < ().05 level. ***Logistic regression parameter estimate statistical y significant at p <0.0 I level. odds ratio, denoted by W, is defined as the ratio of the odds for X = 1 to the odds for X = O,given by * = A/Et, where, A=m(l)/[(1 –m(l)] and B = T(O)/ [(1 - T(O)]. In simple terms, an odds ratio of two implies that when X = 1 the outcome (event) is twice as likely, while an odds ratio of 0.5 would suggest the event is only half as likely to occur. Computed odds ratios are in table 7. Compared to the other two groups, the odds were higher that members of the Low Finance Priority cluster would have a much larger change in perception of the importance of financial management, tend to have more than 10 years of farm management ex- perience, be less satisfied with the workshop ex- perience, and view workshop skills as contributing strongly to net worth growth. For example, the odds were greater than 1.0 (1.46) that a member of the Low Finance Priority group would have more than 10 years of farm management experience. The odds were only about 0.4 that a member of the Low Finance Priority group would end the workshop moderately or highly satisfied with the learning experience, which suggests that they were more likely to not be in these categories. Members of the Low Finance Knowledge group were more likely to have a large change in the knowledge level of financial statements and plans, respectively 4.35 and 2,72. Members of this cluster also tended to have less farm management experi- ence, more off-farm income, and to be less satis- fied with the workshop experience, Members of both the Low Finance Priority and Low Finance Knowledge clusters tended to find the workshop less satisfying than the High Finance Knowledge cluster. Thus, we surmise that the lower the finance knowledge and priority, the harder the producers had to work to master the finance concepts, and the Parsons et al. A Financial Training Program 249 more unfamiliar the topic, the less satisfying the learning experience. Note that the Amish farmers were about twice as likely to be members of the Low Finance Knowledge cluster. The odds were 1,82-to- 1.0 that members of the High Finance clus- ter would have 6–10 years of farm management experience, and that 2.21 -to-l.O members of this cluster would have farm sales greater than $100,000. The odds were about 1.6-to-l that a High Finance Knowledge member would estimate that workshop skills could increase farm net worth gains annually by $ 1,000–$5,000. Thus, logit analysis of the clusters allowed us to identify and understand the differences in impacts of the work- shop and farm characteristics among the clusters, Concluding Comments The borrower training program addressed outreach education issues that are frequently critical to workshop success. In-depth workshops with sev- eral days scheduled for presentations, exercises, and homework are ideal for training on complex topics such as agricultural finance. That stated, the logistics of organizing concurrent workshops re- quires a sharp focus on information presentation efficacy. In our case, the more high-tech satellite up-link approach was not sufficiently flexible to accommodate the scheduling and workshop lead- ership needs of the typical county agent. The draw- back with using pre-taped videos, the alternative, was that careful editing and frequent updating of the tapes were required to accommodate changes made to curriculum text materials. A key finding of this study is that the finance workshops were very successful in terms of knowl- edge gains and potential impacts on net worth growth and farm survival for most participants. The cluster and logit analysis provided some more specific information for subsets of participants, It is noteworthy that the small farm and lower- educated participants benefited relatively more in terms of change in knowledge of financial state- ments and planning than their neighbors with more education and larger farms. Obviously, writing the text at a lower level than most extension materials, emphasizing exercises, and repetition and review were elements of the curriculum that made it ac- cessible to these less-educated participants from smaller farms. However, a key challenge is to de- velop educational approaches that increase the sat- isfaction levels of these less prepared and moti- vated participants, in this case the Low Finance Priority and Low Finance Knowledge producers. Clearly, these two clusters entered the workshop with more deficiencies than the High Finance Knowledge group. This uncomfortable learning challenge needs to be made as positive as possible without lowering the knowledge achievement stan- dards of the course, Two other important impacts from the training experience were that agronomy and dairy science agent site leaders became more knowledgeable of and confident with agricultural finance concepts, to the point where several chose to present the mate- rials themselves rather than to use video-tape pre- sentations. In addition, Cooperative Extension was able to integrate clientele previously not reached by extension programs. Finally, a challenge for Cooperative Extension is to cultivate ties with or- ganizations such as FSA/USDA so that our strengths as educators can be employed with pro- ducers who otherwise would not take the time to master difficult concepts. The borrower training workshops will ultimately enhance USDA and Co- operative Extension partnerships in working with minimum resource producers. References Bernhardt, Kevin J,, John C. Allen, and Glenn A. Helmers, 1996. “Using Cluster Anafysis to Classify Farms for Con- ventional/Alternative Systems Research,” Review ojAgri- cultural Economics 18-4:599-611. Federal Register. 1993. Rules and Regulations. Washington, D.C. 58-259(Dec 30): 16190-16198. Hanson, G.D. 1995. “A Distance Learning Approach to Bor- rower Training.” Agricultural Finance Review 55:133-46. Hanson, G.D, 1997, “The Agricultural Finance Legacy of the New Deal.” The Southern Business and Economic Journal 21:34-46, Hiel, Edwin, and David Herrirrgton. 1997. “Plausible Uses and Limitations of Videoconferencing as a Tool for Achieving Technology Transfer.” Journal of Extension 35(4): 7 pages. (Electronic Publication). Hosmer, D. W., and S, Lemeshow. 1989, Applied Logistic Re- gression. New York: John Wiley, Madalla, G.S. 1983. Limited-Dependent and Qualitative Vari- ables in Econometrics. New York: Cambridge University Press, Peterson, Gary. 1999. Coordinator of Agricultural Short Courses, Penn State University. Personal communication. August 10. SAS/STAT Users Guide. 1989. Version 6, Fourth Edition, Vol- ume 1. Cary, NC: SAS Institute, Inc. 250 October 2000 Agricultural and Resource Economics Review Appendix Table. Logistic Regression Results from 1998-99 FSA Finance Workshops by Cluster Parameter Estimates Low Finance Priority Low Finance High Finance Cluster 1 Knowledge Knowledge Regression Variable (n = 135) Chrster 2 (n = 114) Cluster 3 (n = 1 13) 1. 2. 3. 4. 5. 6. 7. 8. 9, 10, 11, 12. 13. 14, 15. 16. 17. Intercept –1.781 2.581*** -2.619 –1.584*** 1,996 –1,767*w+Change in the view of the importance of farm financial management Change in the knowledge level of farm financial statements –0.640”” 1.470*** -1.241*** Change in the knowledge level of farm financial plans -0.299 1,002*** -J.113*** Farm management experience: 6–10 yrs 0.154 0,376 -0.348 –0.255 0.601 0.001Farm management experience: more than 10 yrs Moderately satisfied with workshop –0.793”” –0.999”” -0.046 -0.341 0.017 0.422 -0.406 -0.345 -0.463 0.273 0.520 -0.398 1.450*** 1,627*** 0.795 -0.123 -0.699” 0.499 Highly satisfied with workshop Farm sales greater than $100,000 Off-farm income: $1-$7500 Off-farm income: more than $7500 Workshop skills will increase farm net worth between $ 1–$5000 Workshop skills will increase farm net worth by more than $5000 Typical profit: $0-$10,000 0.142 0.079 -0,213 0.200 0.084 0.109 -0.414 –0.580* –0.071 0,245 0.471 -0.153 Typical profit: greater than $10,000 Participants not attending any other extension workshops in past year Amish farmer -0.482 0,751* 0.161 *Parameter estimate statistically significant at p <0,10 level, * *parameter estimate statistically significant at P <0.05 level. * **parameter estimate statistically significant at p <0.01 level.