RELATIONSHIP OF DAIRY FARM NET INCOME T0 SPECIFIED FARM MANAGEMENT FACTORS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY JOHN A. SPEICHER 1 9 6 3 mists 0-169 N N “N “I “W N “I“ 11 H “IN 3 1293 11 6186 This is to certify that the thesis entitled RELATIONSHIP OF DAIRY FARM NET INCOME TO SPECIFIED FARM MANAGEMENT FACTORS presented by John A. Speicher has been swepted towards fulfillment of the requirements for _£_b_JL_degree mm /2>7 eéimgdr Major professor Date 3’/ .1 f/é/ LIBRARY Michigan State University 9"»qu av I"? HUNG & 3015' 800K BINDERY INC ueamv muons; MIIEMI.“mlN -‘ 2M OVERDUE FINES ARE 25¢ FER DAY PER ITEM Return to book drop to remove this checkout from your record. RELATIONSHIP CF DAIRY FARR NET IFCCLE TO SPECIFIED FARM KAHAGELEHT FACTORS By John A. Speicher AN ABSTRACT OF A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Dairy 1963 ABSTRACT RELATICNSHIP CF DAIRY FARE NET IICCK? TO SPECIFIED FARM RAEAGELEET FACTORS by John A. Speicher Michigan dairy farms that utilized both Mail-In Farm Account records and D.H.I.A. or Owner-Sampler records for any year in the period from 1958 through 1962 were used as a source of data. A total of 1,041 farm record years was included in the study. A prediction equation was developed based upon a multiple regression model. The degree of curvilinearity of each management factor was established by singularly correlating the factor to net income. The 58 farm management factors were then classi- fied into groups measuring characteristics of the farm en- terprise, and the effects of different combinations of factors within groups were studied as a means of reducing the number of factors to those making a significant (P < °Ol) contribution in the explained variation in net income. Selected factors measuring size, crOp efficiency, livestock efficiency, labor efficiency, costs, intensity and organization were combined to form the prediction equation. The correlation of the 14 management factors ac- cepted as the independent variables in the prediction John A. Speicher equation with net income gave a coefficient of determina- tion (R2) of .75. The average net income was $3,174 with a standard error of estimate of $2,577. Farm management factors which were measures of size exhibited a slight degree of curvilinearity when correlated with net income. The two size factors used were number of cows and number of tillable acres and were found to account for 28 percent of the total computed direct and indirect effect of the factors on net income. Measures of crop efficiency used were crop value per tillable acre, soil value rating, and percent cash crOps. Net income was found to increase at a decreasing rate as any of the crop efficiency factors increased. The combination of crop factors explained 25 percent of the computed direct and indirect effect of the management fac— tors on net income. Crop value per tillable acre was credited with the major portion of this effect. Livestock efficiency factors which were measures of output were linear while those measuring price were curvilinear. Live— stock factors accounted for 29 percent of the total com- puted direct and indirect effect of the factors on net income. Livestock income per $100 feed expense was credited with 75 percent of this effect. The presence of John A. Speicher the income to expense ratio appeared to mask the effect of both level of production and milk price. Kachinery ex— pense per tillable acre and number of tillable acres per cow accounted for 12 percent and 4 percent, respectively, of the computed direct and indirect effect of the factors on net income. When the sample was sorted according to breed, the relationship of the facto*s to net income for Holstein herds was comparable to that reported for the total sample. Jersey and/or Guernsey herds, however, were effected to a greater extent by the size and crop factors and to a lesser extent by the livestock efficiency factors. RELATIONSHIP or DAIRY BARR tar ILCOLE TO SPECIFIED FARE KANAGEQQJT FACTCES By \ John A. Speicher A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Dairy 1965 ACKNO .‘JLED anni s The author wishes to express appreciation to all Inembers of the Department of Dairy for their counsel and ‘understanding during the course of this study. This ex- pression of appreciation is particularly extended to Dr. C. A. Lassiter, Chairman, for the Opportunity and encour- agement to undertake and complete this study, and to Dr. C. E. Meadows, Professor, and Dr. L. D. McGilliard, Pro- fessor, for their many helpful suggestions. A sincere expression of appreciation is extended to the members of the Department of Agricultural Economics for their Willingness to c00perate to such a full extent in all phases of this study. The author is particularly appreci- ative of the assistance furnished by Dr. L. L. Boger, Chairman; Dr. W. H. Vincent, Professor; Dr. L. R. Kyle, Professor; and Dr. J. R. Brake, Associate Professor. A special debt of gratitude is felt for the coopera— tion shown by Mrs. A. M. King, Supervisor in the Department of Agricultural Economics, and Mr. A. J. Thelen, Super- visor in Dairy Herd Improvement Association Incorporated, in providing both counsel and access to their respective farm record programs. The author is especially grateful to his wife Lucille for her unselfishness and understanding throughout the course of this study. ii TABLE OF CCJDDLIS II-ITRODUCDILIN— O O O O O O O O O O O O O 0 REVIEW!- OF LITJJATA TU..IJ£ o e o o o o o o o 0 Description of the Various Farm Hanagement Factors . . . . . . . . . . . . . 7'1 Association Between Farm Incomes and rarm Manageme 1t Factors. . . . . . . . L“? 41114-44; :13. XL ill ALDU L114 0 e o o o o o o 0 Acquisition and Development of the Sample Selection and DevelOpment of Farm ment Factors. . . . . . . . . . . Analytical Design and Lietliod. . . MSUL TS J‘LIuD 318011—5510an o o o o o o o o o APP ‘1:DIX O O O O O O O O O O O O O O O 0 iii Manage— 12 35 55 101 104 109 111 Table Page 1 Relationship of Various Farm Management Factors to Labor Income, 755 Farms, Tomp- kins County, Lew York, 1927. . . . . . . . 14 2 Association of Labor Income and Several Farm Management Factors with Size and.In- tensity of Farming Operation, 1,016 Farm Record Years on Selected Michigan Farms, 1955-19580 0 o o o o o o o o o o o o o o o 24 5 Returns per $100 Feed Fed to Different Classes of Livestock . . . . . . . . . . . 58 4 Relationship of Farm Management Factors to Het Income, Michigan Dairy Farms, 1,041 Farm Record Years, 1958-1962.. a. Measures of Size, Crop Efficiency, Kachinery Cost, and Crganization. . . 70 5 Relationship of Farm Management Factors to Net Income, Michigan Dairy Farms, 1,041 Farm Record Years, 1958—1962. b. Measures of Labor Efficiency and Livestock Efficiency. . . . . . . . . 71 6 Relationship of Farm Management.Factors.to Ret Income, Michigan Dairy Farms, 1,041 Farm Record Years, 1958—1962. . c. Measures of Feed Inputs and Percent Cows in Milk. . . . . . . . . . . . . 72 7 Coefficients of Determination (RC) as Various Livestock Efficiency Factors Jere Added to Dairy Outputs in a Correlation with Net Income. Michigan Dairy Farms, 1,041 Farm Record Years, 1958-1962 . . . . 81 8 Measures of the Effectiveness of Various Farm Management Factors in Influencing Net Income, Michigan Dairy Farms, 1,041 Farm Record Years, 1958—1962. . . . . . . . . . 91 iv Table 10 LIST CF TABLnS—-Continued Statistics Derived from the Prediction Equation of Rot Income Using Farm Ranaje— nent Factors as Variables, hichi;an Dairy Farms, 1,041 Farm Record Years, 1958-1962. Direct and Indirect Effects of Selected Farm Management Factors on Ket Income nhen Eerds were Classified by Breed. Michigan Dairy Farms, 1958-1962 . . a . . . . . \O R) Appendix Table 1 Page r“; b C) S (D l r F- air Avera3e Yearly Values f ment Factors, Lichi :n 1,041 Farm Record Ye' rs a. IMe sures of Size, Crop ulIlCleICJ, Lachinery Costs, ard crsihization . 112 rm - Rim 0 D Avera3e Yearly Values for Farm Lana3e- ment Factors, Lichigan Dairy Farms,. 1,041 Farm Record Years. b. Measures of Lab or Efficiency and Livestock Efficiency. . . . . . . . 115 Average Yearly Values for Farm Lanage— nent Factors, Iichi3an Dairy Farms, 814 Farm Record Years. c. Measures of Feed Efficiency and Percent Cows in Lilk. . . . . . . . 115 Average Acreages, Percentages, and Yields of Crops Grown on COOperating Lichi3an Dairy Farms, 1,041 Farm Record Years, 10 )58—9 ' . . . . . . . . . . . . 116 CrOp Prices and Values as Used to Deter- mine CrOp Value per Tillable lore and Crop Acre Value for Coo peratin3 Licoiban ain F‘arns, 1 ,041 Fa m Record Cbs erva- DIODS. . . . . . o o . . . . . . . o . . 117 Range in Values Observe d for Farm Man— agement Factors. COOperative, Lichigan Dairy Farms, 1,041 Farn Record Years, 1958—1962. . . . . . . . . . . . . . . . 119 Ranges in Values Observed for Farm flan- a3enent Factors. COOperative Lichi; :an Dairy Farms, 1,041 Farm Record Years, 1958—19620 0 o o o o o o o o o o o o o o 120 vi INTRODUCTION The income derived from a dairy farm Operation is the ultimate concern of the dairyman in his role as man— ager of that business. Dairying as practiced in Michigan is basically a means of marketing the products of land and labor. With a system of dairying where the majority of the feed for the dairy herd is produced on the farm, factors such as the crOpping system and crop yields, size and organiza- tion of the farming operation as well as dairy sales and expenses are of importance if maximum income is to be achieved. A tool in analyzing farm Operations and organiza- tion used by workers in the field of farm management is the adaption of certain measures of farm business to a scale or standard to serve as reference points. The rea- soning advanced in defense of this method of analysis is that it has the advantage of speed of analysis and that by comparing a series of farm management standards against a particular farm it is possible to spot the weaknesses and strong points of that farming Operation. This study was designed to study the effect of a number of these measures of farm business activity or farm management factors on the net income derived from the dairy farm. The major objective set forth in carrying out this study was to determine which of these farm management factors were of importance in explaining the variation in net income. The primary means of studying the association of net income with these farm management factors was with multiple correlation analysis. The sources of the finan— cial and production records for this study were the Michigan Mail-In Farm Account Record Project under the direction Of the Department of Agricultural Economics as well as the Dairy herd Improvement Association and Owner- Sampler records under the supervision of the Department of Dairy. The sample used for the analysis consisted of those Michigan dairy farms on which both record systems were utilized for any of the years of 1958 through 1962. REVIEW OF LITERATURE Financial success in dairy farming results from technical Operations involving cows, acres, crOp yields, milk production, feed consumption, and other physical factors. Measures of the technical efficiency in the various farm enterprises have been used in the analysis of farm operations almost since the first attempts to develop and use farm records. Boss (5) in discussing the history of farm cost accounting records stated that in the search for satisfactory measures of cost, emphasis was turned toward determining the physical factors of cost encountered in farm Operation and management. These meas- ures of the man, horse and machine input, and the physical quantities of the elements used in production allowed a determination of the cost on the basis of prevailing wage and price factors. Boss further points out that these physical measurements allowed comparison between farms and farming areas which then spotlighted the most efficient operators and the most profitable practices in production. Description of the Various Farm Management Factorsl Farm management factors have been develOped which measure a large number of the aspects of farm organization and operation. Many of the factors are repetitive in that they measure the same characteristics. This repetition of measurement requires a classification and description of the various factors which have been used before a fuller review can be made as to their influence on farm returns. HOpkins and Heady (8) in discussing farm management fac- tors classified them into groups measuring eight different characteristics of the farm enterprise. This classifica- tion is presented below: 1. Size of Business. Size may be measured in rela- tion to land, labor, capital, or some particular characteristic of the type of farming under study. 1In trying to give a general title to these varied measures of farm activity, farm management workers have suggested and used numerous terms. Some of the titles more commonly used are farm management factors (14,17), farm practices (13), farm efficiency factors (8,14), farm fac— tors (15), factors in success (7), and farm business meas- ures (19). In reporting the findings of the various stu- dies, the term farm management factors will be used as a general title throughout this paper so as to furnish more continuity. Total acres under cultivation. Gross investment measures the value of all inventories involved in the farming opera— tion. This includes the value of the land. humbers of livestock measured in terms of cows, sows, hens, or animal units. An ani- mal unit is used as a common denominator for different types of livestock and is ordi- narily considered to be an amount of live- stock equivalent to one cow, bull, steer, or horse and is based primarily on feed con- sumption. Acres in a selected crop is often used where there is an extremely predominant crOp grown in the locality in question. Labor input as measured by the number of men involved in the farming Operation. Productive man work units (P.M.fl.U.) measure size in terms of the total labor input that would be required if work on the productive enterprises were done with the normal effi— ciency found in the region. This factor re— quires the establishment of normal effi— ciency before it can be used. Total annual input when used as a measure of size includes both fixed and Operating expenses. L‘, Efficiency in CrOpping System. Factors used to examine the cropping program may consider yields, inputs or composition of crops. Fercent of land in a selected crop is an effort to show the effect of growing crops with a higher value per acre or simply to demonstrate the effect of some particular crop. Fertilizer expense per tillable acre is a measure of crop input. Soil productivity rating may also be con- sidered a measure of crOp input as land is an input in crOp production. Crop value per tillable acre measures the results of yield difference as well as the effect of kind of crop grown. Crop yield index has been developed to fur- nish a comparison of crop yields between different farms. The index is basically the relation of the weighted average of the crop yields on a given farm to the weighted average of the crop yields for all farms in a given locality or sample. 5. Livestock Efficiency. The livestock prOgram may be examined in terms of effect of physical output \ and/or input, marketing efficiency, enterprise combination or any combination of the three. Returns per $100 feed fed reflects the de- gree to which livestock is marketed to ad- vantage, the efficiency of physical produc- tion, and the degree to which a profitable combination of enterprises has been attained. Value of animal product is used to reflect both marketing and physical efficiency as it is in essence price times output. Quantity of animal product is a measure of physical output and may be reported as pounds of milk, number of eggs, or rate of gain. E§§d_£§d may be reported as feed used to pro- duce a unit of livestock product, kind or quality of feed, or simply a measure of total nutrient intake. 4. Labor Efficiency. The productivity of labor is ordinarily measured by relating it to the quan- tity of either land, livestock, or both. Crop acres per man is most commonly used where the kinds of crOps grown in the stud- ied locality are reasonably uniform. Livestock production per man has been used in such a manner as to relate the man to I .o-A either the value of livestock output, live- stock numbers, or output in terms of physi— cal quantity for a particular class of live— stock. Work units per man is designed to reflect the influence of both crOps and livestock as it is calculated by dividing the number of productive man work units by the number of men. 5. Machinery Economy. Machinery cost per crOp acre is designed to get at the effect of machinery expenditures. This factor is computed by dividing the total annual cost of power and crOp equipment, in— cluding repairs, fuel, depreciation, and equipment supplies, by the number of crop acres. Cost Ratios. Such ratios are sometimes used when working with a restricted type of farming within a limited locality to determine whether costs are high or low. Costs per acre is a factor used to study to- tal cost or in some instances a particular type of expense such as veterinary expenses. Operating ratio is the percentage which Oper- ating expenses absorb out of the gross income, and is designed to show the proportion of total income used in hiring labor, buying feeds, fuel, and supplies, and in keeping equipment in Operation. It is computed by dividing total Operating expenses by gross income and expressed as a percentage or ratio. 7. Capital Ratios. Ratios involving capital have been develOped in an effort to get at the effects of capital balance. Capital per $100 gross income is intended to show how much capital is required to yield $100 of total income. Capital per man was develOped to express re- source combination. 8. Enterprise Selection. The farm management fac- tors included under this category are those indi- cating the degree of intensity and conversely the extent of diversification. PrOportion of income from a single enterprise is designed to show the importance of any one selected enterprise. Percent income from livestock indicates the relative importance of livestock in the farm— ing Operation for the locality in question. P‘— 10 Diversity index as exemplified by a simple index which shows the number of enterprises that contribute more than 10 percent of the gross income. Black 23 gl. (2) divided farm management factors into five categories as follows: 1. Size of farm business. 2. Balance and diversity of enterprises. 5. Index of crOp yields. 4. Returns per $100 of feed fed to livestock. 5. Efficiency factors. These workers placed factors dealing with enterprise combi- nation as well as percent of a given crOp under balance and diversity of enterprise. Efficiency factors were considered to be all output-input ratios as well as simple ratios be- tween inputs such as crop acres per man. It is soon realized in observing this list of farm management factors and their descriptions that one of sev- eral factors may be chosen to measure any given character- istic of the farm operation. This was demonstrated by Kyle (11) in comparing three measures of size of business with other farm management factors. Records from 599 Indiana farms for the year 1950 were used to establish the correlation between total capital investment, P.M.U.U., and tillable acres and five other farm management factors 11 as well as net farm income.1 The author concluded that for the Indiana farms studied total capital investment was the better measure of size and reduced the variation in the other farm management factors and net farm income more than did the other two measures of size. Productive man work units appeared to be the poorest measure of size. The sample observed in this study was drawn from the entire state of Indiana and was representative of varying soil and climatic conditions as well as different kinds of farming Operations. It appears that in this study total capital investment did the better job of serving as a common denomi- nator and describing the effects of size for all the farms involved. In a study reported by Eisner (12) in 1927, however, seven measures of size were studied using tabular analysis as the analytical tool. When records for 755 farms in Tompkins County, New York, were used to show the associ- ation of these measures of size with labor income, it was observed that a small change in P.K.W.U. was accompanied by a greater change in labor income than with any of the lGross income is defined as total cash income + in- ventory change. Net income is defined as gross income - total cash expenses. Net farm income is defined as net income - value of family labor. Labor income is defined as net farm income — interest on investment at 5 percent. Management income is defined as labor income - value of Operators labor. 12 other measures of size. The number of dairy cows per farm approached P.K.W.U. in this respect. It should be noted that this study occurred in 1927, an entirely different era in time, and over a smaller locality with more uniformity in type of farm as well as in soil and climatic conditions than that reported by Kyle. Association Between Farm Incomes and Farm Management Factors In 1924 Case and Mosher of Illinois (5) examined farm accounts kept by 19 Woodford County farmers. Farms were divided on the basis of the seven high and seven low for each of the five factors examined. The average manage— ment income was computed for each high and low group, and the difference between the two groups was felt to be indica- tive of the effect of the factor on management income. CrOp index, an index of the yields of corn, oats, and wheat, and expense per $100 gross income resulted in differ— ences in management income of over $2,000 whereas results from 3100 invested in productive livestockl furnished a difference of only $1,600. The difference in management income between the high and low groups was less than 31,000 when.the remaining two farm management factors, crop acres worked per man and crOp acres worked per horse, were lProductive livestock was all livestock except horses and mules. 15 examined. This particular study lacked the size of sample or the refinement of analysis to allow very much inference regarding the results. It did, however, represent a pi— oneer attempt at analyzing the farming Operation by the use of farm management factors. The authors felt that the study had additional value in pointing out that the farm which does fairly well in most of the factors stu- cli ed was more likely to be a profitable Operation than the farm that excells in one or two factors and does poorly in others. New York workers (18) were among the first to make use of farm management factors. This early work culmin- ated in the Tompkins County farm management survey in 1927 in which the relationship of a large number of factors to returns were studied. Sample size in this study as reported by I‘viisner (12) was 755 farms. A total of 29 farm management factors was examined by tabular analysis, and for each of the factors the farms were divided into three groups, with the groups being reported by their class limits. The labor income reported was the average for each group. The rela- tionships of the more meaningful farm management factors to labor income are shown in Table l. The tabulations as pre— sented point out several things to be considered. The Table 1 14 Relationship of Various Farm Management Factors to Labor Income, 755 Farms, Tompkins County, New York, 1927.1 Productive Man Work Labor Units Income - — 196 S -12 196 — 514 167 515 - - 572 Acres of Labor Crop Income - - 50 $162 50 - 79 258 80 - - 555 Percent of Cattle Units Labor Heifers Income 1 - 10 3502 11 - 20 256 20 - - 548 Value of Dairy .Products Labor Per COW Income None 3 6 $$ 1 - $ 79 -4 80 — 159 150 140 - - 685 Acres of Crop Labor IPer Man Income _ _ 51 3142 51 -45 190 46 - - 574 lMisner, E. County, New York. 1942. G. Number Total of Labor Capital Labor Cows Income Investment Income --5 $95 - -:’1>‘7,0oo $391 5 — 9 175 $5 7,000-ifl0,w0 245 10 - - 652 filIJXXD- - 425 Percent of Receipts Kan Labor from Labor Equivalent Income Livestock Income - - 1.2 $259 - - 59 $126 1.2 - 1.6 296 60 - 89 580 1.7 - - 207 90 - - 224 Percent of Receipts Pounds of from Labor Milk Sold Labor Crop Income Per Cow Income 1 - 15 3559 None i 91 16 - 55 516 - -5,500 105 56 - - 285 5,500- - 751 Crop Labor P.M.W.U. Labor Index Income Per Man Income - - 96 w 57 - - 15o 5-129 96 - 114 272 150 - 209 178 115 - — 597 210 — — 641 Percent of Receipts Acres of Spent Labor CrOps Per Labor for Labor Income Animal Unit Income - - 46 $896 - - 4 $469 46 - 75 180 4 - 5 222 76 - - -242 6 — - 79 Thirty Years of Farming in Tompkins New York Agr. Expt. Sta., Bull. 782. 15 marked difference between the make up of farms in the era under which the study was made and the present era becomes most obvious as factors such as the number of cows, total capital investment, acres of crOp, and value and pounds of milk are examined. Another point which comes to light in observing the data is the failure of the author to pre- sent the number of farms in each of the groups under the various farm management factors. This lack of information makes it impossible to attach any real meaning to the analysis. At the same time the work does represent a pioneer study and certain inferences can be drawn from it regarding factors of importance in farming for the time and locality in question. Several of the farm management factors have a range in labor income between the low and high groups of over $580, these being P.N.W.U., milk sold per cow, value of dairy products per cow, and P. M. W. U. per man. In addi— tion crop index and number of cows have a $540 difference between the low and high groups. The factors mentioned above are either measures of size of business, labor effi- ciency, or yield factors. Further examination of the fac- tors reveals little difference between groups for factors which measure enterprise combinations. Percent of re- ceipts from livestock and percent of crops are examples of this type of farm management factors. The greatest 16 difference for any factor is that shown for percent of re- ceipts spent for labor. The difference of $1,138 coupled with the observation that man equivalent was associated with no consistent difference in labor income might well lead to the conclusion that under Tompkins County condi— tions in 1927 moving to a two-man farm was of little value in increasing labor income, particularly if this move involved hiring labor. Warren (18), in discussing the Tompkins County sur- vey, states that while the results obtained may not have the mathematical precision required by the laboratory technician, they are sufficiently accurate to lead to ac- curate conclusions. In proceeding with the discussion, he attributes the discovery of the principles of size of business, crOp yield, and production per animal to the Tompkins County survey. Hopkins (7) made use of 325 Iowa Farm Business Records in 1927 and 450 in 1928 in an effort to discover ‘what relationship exists between some of the more common farm management factors and the net farm incomes under Iowa.conditions. An increase in acres in corn was asso- ciated with a corresponding decrease in net farm income up ‘to about 40 acres in corn at which point net farm income increased and followed the increase in acres. Increase in yield.of corn from 20 to 50 bushels was associated with an 17 increase of income of only about $600, while an increase in yield from 50 to 70 bushels resulted in about 51,100 more income. Size of the hog enterprise as measured by the number of sows exhibited diminishing returns in that net farm income rises to 40 sows and then decreases. The author observed that more than 40 sows were in excess of what one man could handle and that the same care for the sows was not forthcoming beyond this point. Percent re- ceipts from livestock as well as equipment expense per crop acre contributed very little to the analysis. An increase in net farm income accompanied an increase in re- turns per $100 feed to approximately $200 returns per $100 of feed, then leveled off and decreased. It was stated that this decrease was due to expenses other than feed increasing. when percent of expense to income was treated as a farm management factor and compared to returns, it was observed that net farm income increased as expense in— creased for a very short range and then decreased rapidly ‘with.an increase in expense. Jhen increases in either Inonths of labor or crop acres per man were associated with :net farm income, income was found to increase at a decreas— ing;rate. Net farm incomes were estimated for the 755 farms included in the study. When the estimated net farm incomes \Nere correlated with the actual incomes, a correlation 18 coefficient of +.85 was obtained from the 1927 figures and +.85 for 1928. The standard deviation of the differences between the actual and the estimated net farm incomes was reduced 46% in 1927 and 44% in 1928. The author does not give the method used to make the estimations of net farm income. Schrumpf (15), in studying potato farms in Aroostook County, Maine, found the relationship of size of business, as measured by bushels of potatoes per acre, to be positive in a year of relative prosperity, but negative in 1928 and 1950 when most farmers sustained losses. In 1928, 98% of the farms having less than 25 acres of potatoes had labor incomes larger than the average of all farms. In compari— son, only 5% of the farms having 55 or more acres of pota- toes per farm returned larger than average labor incomes. In 1929, a year of profit, only 5% of the farms having less than 25 acres of potatoes per farm had a labor income exceeding the average of all farms. Of the farms with 55 or more acres, 86% had larger than average labor incomes. This study over the three-year period, 1928 through 1950, Was based on information collected by personal interviews With 165 farmers. The study also found the yield rate of POtatoes to be significantly related to labor income. In- creased yield of potatoes per acre was associated with de— creased losses in the unprofitable years and increased l9 gains in the profitable year. There was no significant relationship between size of business and yield. A nega- tive correlation was observed between investment in potato machinery per acre and labor income. Small farms had a larger machinery investment per acre than large farmso The difference in crop yields had more to do with placing farms in the different income groups than any other one factor in an examination of farm management factors by Mosher and Case (15). The 57 Illinois farms studied in 1957 were sorted on the basis of crOp yields, and it was observed that the 19 farms in the high group received 3484 more labor income han the 19 farms in the low group. Second only to crOp yields was the efficiency of livestock.1 The high one-third of the farms with re— spect to livestock efficiency received a labor income $589 above the lowest third. When the value of feed fed per acre was tabulated on the same basis as the two farm man- agement factors above, it was accompanied by a difference of $176 in labor income in favor of the high group. In addition to examining the effects of the farm management factors on labor income, the authors also studied the ef- fects of several factors on returns for feed fed to dairy lEfficiency of livestock was defined by the authors as a percentage figure which measures the returns for feed fed weighted according to the amount fed to each kind of livestock. 20 herds. It was found that such returns were influenced about equally by the prOportion of the herds consisting of cows being milked and the amount of milk produced per cow. Herds were grouped so that the high group consisted of 14 herds in which 71.1% of all cattle were milked and which produced an average of 8,815 pounds of milk per cow, while the low group consisted of 14 herds in which only 45.5% of the cattle were cows being milked and which producai an average of only 6,107 pounds of milk per cow. Average returns for the high and low groups were $168 and $107 per $100 worth of feed fed, respectively. The authors failed to define the term percent of all cattle being milked. Thibodeaux 23 g1. (17), in reporting on a study of farm organization in the cotton area of Texas, stated that six major farm management factors accounted for approxi- mately 65% of the variations in labor income on the 157 farms studied during the two-year period, 1951-1952. The approximate effect of each factor on labor income also was determined while simultaneously eliminating any variations in.earnings caused by the other five factors studied. 0n the basis of the relative importance of their effects on farm earning, the factors were classed in the following order: (1) Yield of cotton per acre; (2) percentage of farm land in cotton; (5) returns per $100 of feed fed to .produce livestock; (4) productive man work days per man; 21 (5) size of farm; and (6) number of animal units per 100 acres of farm land. Czarowitz and Bonnen (6), in following up the ear- ‘lier Texas work (17), reported that in the rolling plains area of Texas three factors accounted for 40% of the vari- ations in net farm incomes. Crop yield was reported to account for 21%, acres of crOp land for 15%, and percentage ‘of crOp land in cotton for 4% of the variation in net farm income. This study was conducted in 1955 with a sample .size of 200 farms. The importance of several farm management factors was studied under the conditions of northwestern Indiana by Robertson (14). The study was of 10 years duration, 1929 through 1958, and was based on the financial records of 50 farms. Efficiency in handling livestock was the most important single factor causing differences in labor in- comes among farms. For the lO-year average, the highest one-third of the farms in livestock efficiency had a labor income 3974 above the lowest one-third. Robertson stated that livestock efficiency was largely the result of produc— tion per animal, price of livestock products, and economy of feeding. No mention was made, however, regarding the Inethod of construction of the livestock efficiency index 'used. The importance of crOp yields approached that of livestock efficiency. As the average for crOp yield index 22 mowm.fimm 75 to 126 labor income moved from $276 to $1,192, acfiffmmnce of approximately $900. A positive correlation wasobmnyed between crop yields and both livestock per fflllwflc acre and fertilizer expense. The author reported that large farms had higher laborincomes than small farms, but that the large farms whmfllwere not well managed were the ones with the greatest During the depression years there was a negative losses. It was stated the correlation between size and income. advantage of size was the greater economy in labor, power, The basis for this statement was that the and machinery. farms with the lowest costs for these items per crop acre were larger in size than the other farms. Labor efficiency measured in terms of P.K.fl.U. per man ranged from less than 100 to more than 500, and was correlated with size of farms, differences in equipment, seasonal distribution of labor, .intensity of Operation, and physical and managerial abil— Lflaen tabulated on the basis of P.M.N.U. per man, the ity. .hififll ormw—third showed an average labor income $450 greater than the low one-third. 11 number of farm management factors affecting Rtichjggaxi farms were studied by Jright (20) in a study period of 1955 through 1958 involving 1,016 farm record years. | 1'Farm record years are defined as each farm involved iri tflie’ stnady times the number of years that farm is included. 25 iheefwctof the cropping system was studied by several Iamnmmmgmmnt factors. Farms were divided into three clmmesofswil productivity and it was observed that as soilprmhmtivity increased crOp index and labor income The study showed that as the percent of cash increased. cnnm mnthe better land increased labor income increased, butimatime reverse of this is true on the poorer land; imat alflpmer percent of high value crops increased labor income;eum.that high yields were essential to high labor income. livestock influence on labor income was examined by use of the farm management factors of dairy sales per An farm and productive animal units per tillable acre. these factors was associated increase of either one of with an increase in labor income. Size of business was examined by two factors. The .first of these two size factors, number of cows, showed that time average labor income increased from $588 to $1,2IM5 as iflie number of cows increased from 6.4 for the The effect of the low group to 17.5 for the high group. offlier':aizea.factor, number of tillable acres, is shown in Thais two—way tabular analysis in which the farms Table 2. are divided by size and intensity shows the interrelation- $311133 (31‘ scnne of the problems involved when a sort is made on one or even two farm management factors and then all It is cliaiigxes .111 :income are attributed to these factors. 24 to be observed that as intensity in the form of P.M.N.U. per tillable acre increased labor efficiency, shown by P.M.N.U. per man, also increased for both the large and Table 2 Association of Labor Income and Several Farm Management Factors fiith Size and Intensity of Farming Operation. 1016 Farm Record Years on Selected Michigan Farms, 1955-19381 P.M.W.U. per Number P.K.N.U. Crop Value Productive Till. of Tillable per per A.U. per Labor Acre Farms Acres Man Till. Acre lO T.A. Income (Small Farms--l5O acre average) 2.6 85 107 200 $15.56 1.5 $592 5.5 211 96 221 14.24 1.9 555 4.9 508 82 254 16.67 2.6 751 (Large Farms--26O acre average) 2.4 162 191 251 $12.80 1.6 $688 5.4 167 179 262 14.54 2.2 1,096 4.6 85 172 281 18.54 2.7 1,585 1 Wright, K. T. Dollars and Sense in Farming. iichigan State University Farm Management Pub1., Special Bull. 524. 1941. small farms. CrOp value per tillable and productive ani- mal units per 10 tillable acres also follow the increase in intensity in both size groups. Jith such correlation between factors, the job of attributing causation becomes . I C n I Q . I O A . I . O P O n, .x e O . V C C 9 . . . h a s n o o o o O . O o I A A» 25 extremely difficult, if not impossible, with this type of analysis. The author points out that farmers with small businesses, less than 500 P.M.w.U., had less than one chance in 100 to make a labor income of 52,000, whereas farmers with large businesses, 900 P.M.W.U., had 55 chances in 100 of making a $2,000 labor income. In studying Indiana farm accounting procedures, Kyle (ll) analyzed farm records by several systems to dis— cover significant relationships of various farm management factors with measures of success of the farm business. The years studied were 1950 through 1954 and 1946 through 1950. All the farm records summarized as a part of the Indiana farm record project for these years were used in the study, a total of 6,562 individual farm records. The methods of analysis used were tabular, paired sample,1 and linear correlation analysis. When the relationships of size to net farm income and to labor income were examined by grouping according to tillable acres, a positive rela- tion was shown with the measures of success in 8 of 10 1The method of analysis by paired sample analysis ‘was to first rank all records on one factor such as size. Then each successive pair was split into a low and high half on another factor such as P.M.W.U. per man. The merit of the method over standard tabular was stated to be that interrelationship of variables was reduced. 26 years. In 1951 and 1952 there was a negative relationship. Size was shown to have a more pronounced effect in the high income period 1946 through 1950 than during 1950 through 1954. Paired sample analysis demonstrated a difference of approximately $4,000 in 1948, $2,000 in 1949, and $6,000 inl950 as farms were divided on either tillable acres or P.M.W.U. Net farm income was treated as the dependent variable and each of five measures of size as the independ- ent variable with simple correlation. Total capital in— vestment accounted for 66% of the variation in net farm income, and with each additional $1,000 in investment net farm income increased $161.95. The number of tillable acres accounted for 56% of the variation in net farm income, while number of sows farrowing and total pigs weaned ac- counted for 45 and 44 percent, respectively. P.M.#.U. ac- counted for only 57% of the variation in net farm income and was considered to be a poor measure of size in this study. The function of livestock intensity in the farming operation was studied with the farm management factor, feed fed.per tillable acre. Tabular and paired sample analysis showed little relationship between the factor and net farm income. The multiple linear correlation of tillable acres and feed fed per tillable acre with net farm income re- 'vealed a correlation coefficient of 0.59. Tillable acres l-.—_A 27 alone had a coefficient of 0.55. P.K.N.U. per man, till- able acres in corn, and crOp yield index each exhibited a positive effect on net farm income when analyzed by either tabular or paired sample analysis. The addition of any one of the three factors as a second independent variable of a multiple correlation analysis with tillable acres as the other independent variable raised the multiple corre— lation coefficient from 0.55 to approximately 0.60. A multiple linear correlation analysis was under- taken with net farm income and five farm management fac- tors. A multiple correlation coefficient of .66 was ob- tained and the following prediction was develOped: X1 = + 25.65 X 45.65 X2 + 80.72 X + 9.77 X4 + 117.82 X 5 5 7 - 14,971.56. The variables in the equation were listed as: X1 = net farm income; X2 = tillable acres; X5 = crop yield index; X4 = P.M.N.U./man; X5 = percent tillable acres in corn; and X7 = feed fed/tillable acres. The partial cor- relation coefficients were computed for each of the above variables to determine the relationship of each with net farm income while eliminating the effects of the other four variables. The coefficients are as follows: tillable acres = .7667; crOp yield index = .2951; P.M.W.U. per man = .1796; percent tillable acres in corn = .2589; and feed fed per tillable acre = .1450. 28 Association of nine farm management factors and labor income was studied by Wilkes (19) using 1952 data from 124 South Central Michigan farms. The mean labor income for the farms involved was $5,590 with a standard deviation of $5,698. Three measures of size were used: P.M.N.U., tillable acres, and number of cows. The author stated that of these three only P.M.W.U. proved reliable in estimating labor income. The correlation coefficient indicated that changes in I;M.fl.U. were associated with only 25% of the variation in labor income. The author further stated that for the purpose of estimating either the level of or change in labor income, the three farm management factors dealing with size were poor, with the exception of P.M.J.U. which had some value. The simple correlation coefficients were referred to as the percent of explained variation in income asso- ciated with each of the six remaining farm management fac- tors and were as follows: Income per $100 expense = 74.0%; livestock income per tillable acre = 17.5%; P.M.N.U. per man 16.4%; dairy sales per cow = 14.9%; crOp yield in- dex = 8.4%; and tillable acres per animal unit = 0.1%. Since changes in the level of the farm management factors are often advocated as a means of increasing incomes, the efficiency with which changes in labor income can be pre- dicted from changes in these measures was investigated. Estimating changes in labor income from changes in income 29 per $100 of expense proved to be effective, the regression coefficient being $95.64 with a standard error of $5.02. The regression coefficients for the other efficiency measures also had low standard errors. Management factors influencing the percent return on capital investment for 59 Los Angeles County commercial dairies were studied by Albright (l).- The dairies studied purchased all or most of their herd replacements, ranged in herd size from 141 to 659 cows, handled an average of 65 cows per man, and had an investment per cow figure rang— ing from $665 to $1,640. The production levels of the herds varied from 11,619 pounds of milk and 591 pounds of fat to 15,895 pounds of milk and 547 pounds of fat, with an average yearly concentrate consumption of 5,620 pounds per cow. Standard partial regression coefficients were calculated on a within—year basis for the data compiled from 1956 through 1960. The most important management factor for the study was production cost on a per cow basis. Feed costs, roughage and concentrate, were signifi- cant at the .05 level of probability 5 out of the 5 years studied; labor costs 2 out of 5 years; and the cost of herd replacements and operating costs were significant 5 out of 5 years studied. Prices received for butterfat and the pounds of butterfat produced were significant at the .05 level of probability 4 out of the 5 years studied. 50 The farm management factors of cows per man, hours per cow, percent cows dry, milk produced per cow, culling rate, in- vestment per cow, feeding efficiency, number of cows per herd, prices paid for hay and concentrate, and interest on assets failed to show a significant relationship at the .05 level of probability. Numerous farm record projects are carried on under the direction of the various agricultural economics depart- ments of land grant institutions in the United States. In some instances these results are summarized and the summary published as a part of the prOgram which serves as an edu— cational tool, a research program, and an in—service train- ing situation for the personnel involved. The published report tends to be descriptive in scope, however, as its prime purpose is to point out to farmers some of the items which will be of interest and of aid to them. Examples of such reports are the 1962 Summary of Illinois Farm Busi- ness Records (16), which reports on all types of farming and is general in nature, and the report Dairy Farming Today-~Southern dichigan (9), which has been restricted to one type of farming and in a restricted area. Kelsey and Brown (9), in reporting on Southern .Michigan dairying, divided the 254 dairy farms into three groups according to total investment. within each of the size groups, the farms were divided into two groups on the 51 basis of gross income per $100 total cost. The average of the farms in the high income-cost ratio group had more cows, tillable acres, and gross income than the low income- cost ratio group. The authors noted that there did not appear to be any consistent differences in total invest- ment and tillable acres per man, but gross income and number of cows per man were greater for the groups with over $100 income per $100 expense. The well-balanced farms achieved higher production per cow and consequently higher product sales per cow. It was felt that the fact that labor income increased on well-balanced farms as size increased and that it decreased on poorly balanced farms as size increased was of particular significance. This type of descriptive analysis as used by farm management workers is of particular benefit in working with dairymen but does not lend itself to such a study as this one. The number one thing derived from this review of farm management factors is the difference resulting from various kinds of farms, localities, and periods of time in which the studies were undertaken. The relative importance of most any farm management factor studied has been de— pendent upon type of farms studied, date at which the in- formation was collected and the locality from which the I I farms were drawn. 32 It is felt that a study involving Michigan dairy farms operating under current conditions can be of particu- lar significance to the Michigan dairy industry by deter- mining the relative importance of the various farm manage- ment factors in the profitable dairy farm operation. It, likewise, is not out of line to assume that these relation— ships will be ofvsome merit in localities where dairying is practiced in a similar manner to that practiced in Michigan. EXPEHIHHNTAL EROCEDURE I. Acquisition and DeveIOpment of the Sample The development of high speed computers capable of interpreting and processing large masses of data has opened the way for centralization in the processing of farm record programs. A movement has been made from dispersed centers using hand computation methods to a central location where high speed handling and computing equipment can be utilized. This shift has been coupled with a growth in the breadth and depth of data being sum— marized, analyzed, and returned to the participating farmers in the form of usable information. In the process of handling and working with these farm records, a large concentration of information concerning farm Operations has resulted This study, as mentioned earlier, utilized two such sources of data: the Michigan Dairy Herd Improvement As- sociation records and the Michigan Mail-In Farm Record Pro— ject. Dairy Herd Improvement Association (D.H.I.A.) rec- ords and Owner-Sampler records are dairy production records which are processed as a part of the Michigan Dairy Herd Im- provement Association record system. The D.H.I.A.-I.B.M. _ 33 - 54 records, as the machine tabulated records were first known, came into general usage for all herds on D.H.I.A. for the testing year of 1958. The D.H.I.A.-I.B.M. program is a continuation of the D.H.I.A. records which were previously calculated by the local supervisor. Owner—Sampler records were first machine tabulated in 1959. The hail-In Pro- ject is a continuation and up—dating of the Michigan Farm Record Project initiated in 1928. The pregram of farm financial records started as a pilot project in two coun— ties as a machine calculated project in 1955, at which time the words Mail—In were added to the title and, in fact, became the working title of the project. The suc- cess of the pilot project led to the conversion of the en- tire program to this basis in 1957. An alphabetical card listing of all cooperators on the Mail-In Project was obtained, by county and year, for the years 1958 through 1962. The listing obtained repre— sented approximately 1,600 different Michigan farmers for the year 1958 and 1,200 farmers for each year after that. A listing according to farm number, and county, was obtained for all herds completing a testing year on D.H.I.A. for any year during the 5—year period. The years completed, along with name and address were on this list. Approximately 2,000 herds were included each year on the D.H.I.A. listing. It was not possible to obtain a listing 55 of herds on Owner-Sampler testing; and it was therefore necessary to go to c0pies of monthly reports to find the herds that had completed any of the years studied. Herd numbers on Owner-Sampler testing ranged from 800 in 1959 to 1,275 in 1962. The two systems of records both Operate by assign— ing a farm or herd number to each cOOperator. Numbers as- signed are permanent and are removed from circulation should the cooperator leave the project. The assigned numbers, however, are not common for both sets of records. It was, therefore, necessary to hand-match the farms which were common to both sets of records. Farms were matched as regards name and address of each c00perator. Any co- Operator which appeared to be the same individual by vir- tue of the same mailing address and surname but not identi- cal in given name was checked by referring to individuals who were personally acquainted with the COOperator in question. A further check to establish that names from one record were the same as those obtained from the second record was undertaken later in the study. Data common to both sets of records were matched to determine any gross dissimilarities. The following data were found to be com- mon to both systems: average number of cows, average milk produced per cow as compared to average milk sold per cow, total milk produced compared to total milk sold, value Of product contrasted to dairy product sales, milk price, and breed Of cattle. Dissimilarities were followed up by means Of personal consultation with individuals acquainted with the cOOperators. In no instance was a farm discarded on the basis Of difference alone. A total of 448 farms and 1,404 farm record years was found to be common to both sets of records. These 1,404 observations were com- plete as regards all data normally included by either record system. The production records, D.H.I.A. and Owner-Sampler, are Operated on a testing year which begins on October 1 and ends on September 30 of the following year, while the accounting year used under tre Hail-In project is from January 1 through December 51 Of the same year. In order for the two sets of records to be comparable they must cover the same time span. A January to January testing year was computed for the production records by Obtain- ing the monthly production cards used to calculate the original annual summary and then to re-sort and summarize these cards. The data reported by either type Of record then represents information from the same farms and over exactly the same period of time. In the process Of shifting the testing year from an October beginning to a January beginning, the problem 57 Of bi-monthly tests presented itself. As is occasionally the case with D.H.I.A. and Owner—Sampler records, months missed by the tester for any one Of several reasons are usually covered by reporting days for a two-month period on the following month. The month missed is then computed by using milk weights and fat tests from both the proceed- ing month and the following month. Days from the forepart T# i g--.- Of the missed month are multiplied by milk weights from the preceeding month while days from the latter part of the missed month by weights from the following month. In reporting the information to the herd owner, however, the values are reported for a two—month period. This necessi- tated allotting apprOpriate values to December and January for any bi-months reported in January. In addition the decision had been reached to study the effect Of seasonal milk production on income and this prohibited bi-monthly reports in August. TO furnish greater flexibility in the use Of the production records, all bi—monthly values were divided and allotted to the two months concerned. The method of separating bi-monthly values for milk, fat, feed, and value of product was to divide each value by the total number of days in the two months and then multi- ply by the number Of days in each month. The exception to this was in the case Of certain feed factors, such as ‘ days on pasture, where it was Obvious that the entire value 58 given was for one or the other Of the two months. A total Of 109 bi-monthly records were divided between the appro- priate months. It was desired that the effect of the breed of cattle on net income as well as its effect on several Of the farm management factors to be included in the study be considered. TO study the effect of breed, it was first necessary to classify the herds in question according to breed. Both the accounting records and the D.H.I.A. and Owner—Sampler list the breed of cattle for each OOOperator. D.H.I.A. records, however, are essentially individual cow records and furnish much more detailed information on the make-up of the breed of cattle in the herd. In particular D.H.I.A. records furnished the breed Of any cow in the herd which had finished a BOB-day lactation record. A study Of D.H.I.A. records revealed several features re- garding herd make-up. In many instances the breed re- ported for the herd simply designated the breed in the majority in the herd. It was also noted that the breed make-up within a herd Often changed over a five-year period but the breed designation for the herd did not always fol- low this change. These two features Of breed classifica- tion, along with the occasional error made as the informa- tion was punched into the data cards, made it necessary to devise a more accurate and realistic assignment Of breed Of dairy cattle. Due to a restriction in numbers it became Obvious that the study should be restricted to the Holstein, Jersey, and Guernsey breeds. The few cases involving other breeds were removed from the study. All herds on D.H.I.A. were then examined in detail and classified ac- cording tO breed on the basis Of the breed reported for individual cows within the herd which had completed BOB—day lactations. A herd was accepted as Holstein or Jersey and/or Guernsey if 90% or more Of the EOE-day lactation records were completed by Holstein or Jersey and/or Guernsey cows, respectively. It is commonly accepted that a difference exists between the Holstein breed and the two smaller breeds in regard to both pounds Of milk produced per cow and per- cent Of milk fat. With this in mind, a decision was made to classify according to breed on the basis Of the aver- age yearly fat test. Classification according to milk produced per cow was ruled out as level of production was one of the factors to be analyzed. If milk production had been used as a basis of classification, several high pro- ducing Jersey-Guernsey herds as well as several low pro- ducing Holstein herds would have fallen into the catagory Of mixed breeds and consequently would have been discarded from the study. In this study involving the effect Of yield on income the extremes are Of particular interest. .. B“ “—Il 40 Studies by fiunder (22) indicate a mean fat test for 0/ Michigan Holsteins of 5.6 m with a standard deviation Of .40. The average fat test on Guernsey and Jersey records studied was 4.84 and 5.58%, respectively. The respective standard deviations were .48 and .51. In studying 8,658 Michigan Holstein records, Burdick (4) found the average fat test to be 5.65% with a standard deviation of .40. 771 The finding of these workers indicates that with a ‘F m. . _ herd size Of eight, the smallest in this study, a standard deviation Of .14 and .17 can be expected for the Holstein and Guernsey breeds, respectively. It was thus expected that 99% Of the Holstein herds would have an average yearly test between 5.24 and 4.08%. In a like manner, 99% Of the Guernsey herds would be expected to have a yearly milk fat test of between 4.55 and 5.55 Percent. It then follows that any herd having between 4.08 and 4.55 percent milk fat should be considered as a mixed breed herd. When the average herd size of 59 cows was used in deter— mining the standard deviations, the area between 5 stand— ard deviations in fat test Of Holsteins and Guernseys was from 5.87 to 4.60 percent. Examination Of the fat test for the herds Of known breed which were used in this study revealed that the test for Holstein herds ranged from 5.20% to 4.25% and that the fat tests for the Jersey— ' Guernsey herds ranged from 4.40% to 5.75% milk fat. 41 All herds with an average yearly milk fat test which fell between 4.25 and 4.52% were considered as mixed herds and excluded for this study. The basis for this exclusion was the studies Of the above authors (4,21) coupled with Observation of fat tests from herds which were a part Of this study and for which the breed make-up was known. As a result Of breed classification, 21 dif- ferent herds and 86 farm record years were excluded from this study. This number includes both the herds rejected as minor breeds as well as those rejected for being a mixed herd. Previous studies have demonstrated the necessity Of unifying type Of farm in such a study as this one. TO Obtain this uniformity, the farms were classified by source Of income. The percents Of crop, dairy, beef, swine, poultry, and Off-farm incomes were computed by dividing gross income into that portion Of gross income derived from each Of the above mentioned sources. Dairy farms were considered to be those from which 70% or more of the gross income was derived from the dairy enterprise, either through sale Of dairy products or dairy animals. Farms on which less than 70% Of the gross income was derived from dairy and with only one other major source Of income were classified as dairy-crOp, dairy—beef, dairy-swine, or dairy-poultry farms. General farms were those with several sources of income, none Of which served to F._; “‘ “__1 ‘ffl 42 contribute a majority of the income. Part-time farms were those with 20% or more of the gross income derived from Off-farm sources. A total Of 540 farms with 1,041 farm record years were classified as dairy farms. II. Selection and DevelOpment of Farm Management Factors “FA The premise that the income derived from a dairy farm Operation is the ultimate concern of the dairyman in his role as manager Of that business has been accepted as a part Of this paper. The assumption has been made that the manager desires to maximize his net income and that his success as a manager may be measured by his ability to dO so. The measure Of success used in this study was net income. Net income may be defined as gross income minus gross expense plus or minus the change in inventory. This value was computed in the Mail—In Account records by summing net cash income and inventory change, where net cash income was equal to total cash receipts minus total cash expense. Alternate choices as measures to reflect the degree Of success were net farm income, which deducts a charge for family labor, and labor income, which places a charge on both family labor and capital investment. Net income was selected for its capacity to reflect the earning 43 ability of the farm as a business unit, and because of the fact that net income is a more familiar term that is much more readily understood outside of the field of agricul- tural economics than either of the other two measures of income. Several measures of size of business were consid— ered and the following were accepted: total investment, tillable acres, number of cows, and number of men. The total investment figure used was the average of the begin- ning and ending total investment for the year in question. The total farm investment was considered to be the inven- tory value of land, buildings other than the farm dwelling, machinery, livestock, supplies, and feed on hand on Janu— ary 1. Values used for machinery and buildings were based on the cost less the depreciation claimed for income tax. t/A Land was priced on a bare land basis and represented a conservative market value for the particular land in ques- tion. Livestock values were estimates based on current livestock prices with purchased animals valued at their purchase price. Feed and supplies were represented by market costs. It should be understood that the investment figures used were those reported by COOperating farmers and the only computation needed to prepare the factor for this study was to change the beginning total investment to I an average figure. The calculation of average total an investment consisted of summing beginning total investment and ending total investment and dividing by two. The selection of total investment as one of the farm management factors was as a consequence of the ability of the factor to measure size across differences in enter- prise organization, land quality, and machinery input. An investment figure makes ure of the dollar as a common de- nominator to reduce all inputs so that they may be summed. The acceptance of only those farms that received 70% of their income from the dairy enterprise would lead one to expect that any factor directly related to this enterprise would have a marked influence on returns. The number of dairy cows was then chosen for its expected ability to re- flect the effect of size on a study in which all farms are dairy farms. Cow numbers were also considered important as a factor because of the emphasis currently placed on herd size by workers in the field of farm management and because of the continuing trend to larger herds. The production records were accepted over the Mail- In records as the source of cow numbers because of a more systematic method of determining numbers. The number of cows was calculated as the number of cow days for the year divided by the number of days in the year, where a cow—day was equal to one cow in the herd for one day. Cow days are counted on all animals which have freshened but which are 45 now dry as well as all cows in milk. Cow numbers used in the Mail-In records were those reported by the dairymen. COOperators in the Mail-In project are advised to report the number arrived at through the use of D.H.I.A. and Owner-Sampler records whenever possible. The number of tillable acres was used as another measure of size, in that acres are indicative of the size of the crOp or feed enterprise. The inherent productivity is not the same on all tillable acres, but land does rep- resent the first physical input in a system of farming which used dairy animals as the prime market for crOps produced. Included under tillable acres was land in har- vested crOps, tillable pasture, land devoted to crOps that failed, tillable land reserved for government programs and idle tillable land. The use of total acres was re- jected due to the large amount of unproductive land included in such a figure and because of the unequal percentage of this unproductive land between different farms. The number of men was used in an effort to obtain a factor reflecting the total labor input of the farms studied and the effect of labor input on returns. The source was the Mail-In records and was calculated as a part of these records by summing the number of months of Operators labor, hired labor, and family labor then di- viding by 12. 46 Other measures of size were considered unsatisfac- tory for this study. A productive man work unit or produc- tive day of work represents the accomplishments of a man working at average efficiency for a lO-hour day. The P.M.W.U. on a farm is the estimated total number of 10- hour days of work required to care for the amount of crops and livestock kept. This farm management factor was a part of the information reported by the Hail-In F 2‘. "T "_ project until 1961 at which time it was discontinued. The computation of P.M.N.U. for 1961 and 1962 was not under- taken as a part of this study. In the mechanization of agriculture, the values previously used for the calcula- tion of P.M.W.U. have become unrealistic. Determination of values reflecting the organization of present day agri- culture was beyond the sc0pe of this study. The same logic which prevented the use of P.M.W.U. also prevented the use of total animal units as a measure of size. Returns resulting from crOpping practices can be eXpected to be affected by crOp yields, type of crOp grown, prices received, land use, and crop disposal. Crop value per tillable acre was a factor obtained directly from the Mail-In records. The factor was computed as a part of those records by dividing total crOp value by the number Of tillable acres. Total crOp value was obtained by 4. i 47 multiplying crop production by November prices. These prices are reported in the appendix. CrOp value per till— able acre reflects crOp yields, the market situation, rela- tive value of crOps grown, and land use. The major weak- ness of the factor was considered to be the large number of factors influencing it. It was felt that a number of farm management factors studying the effect of the crOp- ping system on income might well be advantageous. To establish the effect of level of crop yields on income, an index showing the relative productiveness of crops was computed. This crop yield index was calculated for the 1,518 farm record years that were a part of this study after the farms had been classified according to breed. The steps in the computation of such an index were as follows: (1) Determination of the average yield of each crop across all farms in the study, by year; (2) the acre index for each crop on each farm computed by dividing crop production per crOp by average yield for that crop; (5) sum all acre index values per farm; (4) divide total acre index by total acres on the farm used to compute the index; and (5) multiply by 100. These computations may be algebraically expressed as follows: I CrOp yield index = Z (Z; . ai) Y. 1 gth bl 1:: x 100 48 Where y the yield on a "given" farm Y the average yield of the farms used as a base of comparison i = the "ith" crOp a = acres on the "given" farm CrOp yield index was an expression of average crOp yields on a weighted basis. - _-_.._ The distance above or below a crOp yield index of 100 that t J a particular farm may be was indicative of level of produc- tion of crOps on that farm. Values above 100 indicated production superior to the production of the average farm. Crop acre value was designed to account for the vari- ations in crop value per tillable acre resulting from the kind of crOps grown. It was realized that some crops are relatively higher in value than others. Value constants were develOped to reflect this difference in value between crops. The value constant for any crop is the Eovember price for that crop times the average yield of the crOp for all the farms in the study. In calculating the crop acre value for an individual farm the number of acres for each crop grown was multiplied by its respective value con- stant. The products of acres times value constants were then summed and divided by the total number of acres used to produce these crOps. The resulting quotient was the ' crop acre value. Crop acre value differs from crOp value per tillable acre in that the effect of yield and degree of 49 land utilization have been removed. The removal of the yield effect was accomplished when the November price for each crOp was multiplied by average yield for the crOp rather than by the actual yield. The effect of land utili- zation was removed by dividing by the total acres in crOp rather than by total tillable acres. An algebraic expres— sion of the factor is as follows: i ,g'ui . Pi) . a. ' l .5“: the average yield of the farms used as a base CrOp acre value = where Y of comparison P = the November price i = the "ith" crop a = the acres on a "given" farm The inherent productivity of the soil has been ac- cepted as a major factor in effecting crOp yields and to a lesser extent type of crops grown. The farm management factor, soil value rating, represents an attempt to measure the variation in income resulting from difference in the inherent productivity of the soil. The value was obtained from the hail—In Account records and is expressed as a dol— lar value. In developing this rating, county personnel classified the farms within the individual counties on the basis of soil productivity. The classification ran from A through D, with A ranked farms being the best in that par- ticular county. A dollar value was then assigned by county, F;““T for each of the soil productivity classifications. The values assigned for one county may be entirely different from those for another county in a different part of the State. It was this dollar value that has been used as the soil value rating. The weaknesses to be expected in the factor are first those due to errors of judgement in classi- fying soil. Another problem in such a method of classifi— cation of soil productivity is the difficulties involved in attempting to remove value imparted by uses other than agricultural. Closeness to urban or industrial areas are examples of conditions imparting additional value to land over that imparted by agricultural uses. Fertilizer and lime expense per tillable acre repre— sents the use of the dollar value of a crOp input to predict net income. As this factor was selected it was realized that any effect on net income would be contingent upon a large number of other variables. Fertilizer and lime ex- pense per tillable acre has been a widely used factor in‘dm analysis of farm organizations by farm management workers. Resources, such as land, that were not used to their full capacity would be expected to result in an income which was less than maximum. On the basis of this lOgic, the per- centage of tillable acres idle was taken from the Nail-In Account records and used as a farm management factor. The percentage of cash crOps was develOped as a meas- ure of crop disposal. The value of the crops which were 51 considered to be cash crOps were summed. Cash crOps were considered to be the following: wheat, dry beans, sugar beets, potatoes, legume seed, flax, truck crOps, fruit, berries, canning peas and beans, crops sold standing, in- surance payments on crOp damage, and payments through the soil bank program. The value of crops grown as cash crops was added to sales of feed crOps. The sum was then divided 5? by total crop value to determine the percentage of cash crops. It was realized that this factor would be limited in value to the study in that the farms had previously been sorted to remove any farms which had less than 70% of their gross income from the dairy enterprise. It was felt, how— ever, that the factor might account for enough variation in income in herds classed as dairy herd to justify its use. A large number of farm management factors which deal with dairy efficiency have been included in this study. These dairy factors have been justified on the basis that the income of a sample such as the one used in this study could be expected to be markedly influenced by the dif— ferent aspects of the dairy Operation. Dairy sales per cow might be expected to have much the same relationship to the dairy enterprise as might be expected between crop value per tillable acre and the crOp enterprise. Dairy sales are influenced directly by pounds a of milk sold, percent of milk fat, and price received. Mail-In accounts report dairy sales per cow, just as 52 D.H.I.A. records list the Value of product per cow. Both measures are intended to represent the same factor; however, the figure reported by the financial records is one of in- come received divided by the number of cows, while the fig- ure reported by D.H.I.A. is a calculated figure and as— sumes all milk produced has a value. The values used for this study were taken from the Nail—In accounts. Milk production per cow has been studied in more de— tail than the other farm management factors due to the tra- ditional emphasis placed on this measure of physical output by workers in the field of dairying. Milk produced per cow was taken from the D.H.I.A. and Owner-Sampler records and was based, as is the case with production records, on mondfly weighing of the milk produced by the individual cows in the herd. Milk sold per cow was obtained from the Mail—In ac- counts where it was calculated as a part of those records by dividing pounds of milk reported sold by the average number of cows. A problem presented with this factor was the failure to have the pounds of product reported in 1.7% of the cases. The absence of values was circumvented by subtracting the average difference between milk produced per cow and milk sold per cow from average pounds of milk produced. A difference of 679 pounds was found between these two measures of milk per cow. The difference was obtained by subtracting the averages of all observations for milk sold per cow and milk produced per cow. Milk 53 sold was then considered to be 679 pounds below milk pro- duced in those cases where it had not been reported. The two milk factors were converted to the factors of 4% fat corrected milk (POM) produced per cow and 4% FCM sold per cow in an effort to determine if a comparison could be made across breeds of cattle. The multiplication of average pounds of milk by 0.4 plus average pounds of a“ milk fat times 15 resulted in the pounds of 4% ECU. The q fat test obtained from the production records was used in the computation of 4% FOR sold. The factors, milk produced per cow, milk sold per cow, 4% FCM produced per cow, and 4% FCM sold per cow were considered to be four measures of the same factor. It was felt, however, that a comparison of production ability was justified. It should be pointed out that both of the sales factors were dependent upon the pro- duction records either for missing values or for fat tests. Pounds of milk fat per cow were obtained from the production records and were included because of past, if not present, emphasis by dairy scientists on this yield factor. Two milk price factors were used. Milk price per cwt. was obtained from the financial records and in actu- ality was calculated by dividing total dairy sales by -ak total pounds of milk sold. This figure is representative w of the price received after the deduction of marketing expenses, hauling included. As with milk sold per cow, 54 the milk price per cwt. for those few herds not reporting pounds of milk is a calculated figur which is contingent upon the existence of production records. The second price factor used was 4% POM price per cwt. The FCM price was developed from the above milk price and represents the price which would have been received if the milk sold was the pounds of 4% FOE sold per cow as computed pre- viously. Computation of the factor consisted of obtaining , the difference between the fat test from .04 and multiply— ing the difference by the average price on fat differential paid for the year. This product was then algebraically added to milk price per cwt. Price received on fat differ- ential for the period of time in question was as follows and represents the increase in milk price per cwt. for a 0.1% increase in fat test: 1958 = 6.58c; 1959 = 6.71¢; 1960 6.67¢; 1961 = 7.00¢; and 1962 = 6.65¢. The factor of FCM price was develOped under the assumption that when FCM sold or FCM produced was used it also would be used. The majority of the farms in the study sold milk under some type of base-surplus or base incentive plan. Percent base milk was developed to study the effect of seasonal mi k production on income. The production records were used to obtain the total amount of milk produced from August 1 through December 51. The milk produced during , this 5—month period was divided by the total yearly production to furnish the percentage of milk produced dur- ing the base period. The farm management factor dairy cattle income per cow was obtained from the financial records and is the difference in the beginning and ending dairy cattle in- ventories plus dairy cattle sales minus dairy cattle pur- chases. It was felt that there might well be major dif- ferences in income as a result of cattle sales, particu- larly in as much as a number of c00perators were purebred breeders and as such have traditionally been assumed to obtain a sizeable portion of their income from the sale of breeding stock. Since inventory change and purchases were included in the computation of the factor, it was possible to have a negative dairy cattle income. A factor was developed to study the effect that the number of young stock carried might have upon income. The replacement stock to cow ratio was calculated by summing the average number of heifers and the average number of calves, as determined from beginning and ending inven- tories, and dividing by the average number of cows. Both replacement and cow numbers were obtained from the Mail—In accounts. A point to be remembered with such a factor as this is that the replacement to cow ratio may depend more on the percent of heifers calves born than on dairy herd organization. 56 A study of the effect of dairy cattle housing on income was desired. The closest measure that could be obtained was improvement cost per cow. The inventory value of building and improvements, which represents building cost minus depreciation, was divided by the num- ber of dairy cows. The major weakness was realized to be its inability to actually measure differences in types and extent of housing. It was assumed that idle resources in the form of land would have a depressing effect on income, so it was reasoned that idle resources in the form of dairy cows would have a depressing effect on income. Percent of cows in milk was felt to be a measure of the productivity of resources. The factor was obtained from the D.H.I.A. records and was calculated as a part Of those records by dividing total days in milk by total cow days. The number of tillable acres per cow was designed as a measure of the intensity of the dairy Operation. The number of tillable acres as reported by the Mail-In ac— counts was divided by the number of cows to furnish the value for the factor. A number of feed input factors were computed for this study. Feed input as determined by Mail-In records is the value of the feed crops produced plus feed purchases minus feed crop sales plus or minus the change in the 57 inventory value of those feed crops. Livestock income per $100 feed expense was based upon this calculat’on and was computed by dividing total livestock income by feed expense and multiplying by 100. Livestock income is the ’ifference between beginning and ending inventory Value of livestock plus livestock sales minus livestock purchases plus dairy product and egg sales. The factor of livestock income per $100 feed expense was accepted as a measure of the returns above feed cost for the livestock enterprise. An effort was made to determine feed cost per cow through the use of the Mail—In records. Since feed quanti- ties as measured by the financial records were allotted to all livestock on the farm, it was necessary to subtract that supposedly fed to livestock other than dairy cows. Thezmnmod used in an attempt to remove the effect of other livestock was to divide the income for each type of livestock by the returns per $1 feed fed to the various classes of livestock for the year in question. The resulting figures then repre- sented feed expense for each class of livestock and when sum- med represented total non—dairy feed expense. The values of the returns per $1 feed fed to different classes of livestock were obtained from Illinois data (16). These particular values, as shown in Table 5, were chosen because of their completeness. Feed cost figures used for heifers and calves were $105 and $65, respectively, and were obtained by multi— pkfing feai required by local prices. The total costcfi.fimdfhr {ITI' 58 Table 5 Return per $100 Feed Fed to Different Classes of Livestock1 Feef hative Cow Feeder Sheep 'Year Herds Cattle Hogs Raised Poultry 4:3 2‘2? ‘5 .9 15 1958 1.62 1.44 1.80 0.98 1.42 1959 1.47 1.12 1.14 1.02 1.25 1960 1.29 1.77 1.64 1.08 1.57 1961 1.59 1.16 1.64 1.10 1.50 1962 1.49 1.48 1.59 1.26 1.44 1 Summary of Illinois Farm Business Records, 1962. Uni- versity of Illinois, Coop. Ext. Ser. Cir. 874. replacements was then computed using the stated values and subtracted from dairy feed costs. The remaining value was then assigned to the dairy cow herd and when divided by number of cows was listed as the farm management factor, feed cost per cow. The remaining feed factors, as well as the percent cows in milk previously discussed, were taken from the D.H.I.A. records. Owner-Sampler records do not carry feed information and as a result the number of dairy farm record years with these factors was reduced to 814. Average feed cost, as taken from the D.H.I.A. records, was computed by summing the value of grain, hay, silage, and pasture fed. 59 Grain prices used were those reported by the supervisor, and represented market value of the grain mixture. Rough— age prices used were constant for the entire sample and were based on market value. Pasture price used was $6 per month per cow. Three physical measures of feed fed were determined from the D.H.I.A. records: pounds of grain per cow, pounds of hay equivalent per cow per day, and pounds total digestible nutrients (T.D.K.) per cow per day. All three of these factors were determined from values reported by the local supervisors. Grain per cow was based on the weight of the grain ration fed on test day and has been reported as total yearly pounds of grain per cow. Hay equivalent per cow per day is a measure of hay and silage fed during the winter months when all cows were in confine- ment. The factor was computed by dividing total silage by 5 and adding to total pounds of hay reported. Total hay equivalent was then divided by total cow days for the period of winter feeding to furnish pounds hay equivalent per cow per day. The grain quantities were derived from the weighing of grain fed to each individual cow. Rough— age quantities were based on a report listing the average hay and silage consumed per cow daily. Total digestible nutrients per cow per day were obtained by multiplying total grain for the winter feeding 60 period by 0.75 and total hay equivalent for the same period by 0.50. The two products were then summed and di- vided by the total cow days for the Winter feeding period. The factor, TDN per cow per day, was calculated over the Winter feeding period, as was hay equivalent, due to the inability to allocate nutrient intake from pasture feed— ing in a study such as this. Days on pasture was not considered to be a feed factor but rather a farm management factor reflecting dairy herd organization. Calculations of the factor consisted of summing total cow days on pasture and dividing by aver- age number of cows. The number of cows per man, milk sold per man, and 4% F.C.K. sold per man were the three farm management fac- tors develOped as a measure of labor efficiency. The three factors were considered to be an approximate meas— 'ure of the same item. Milk sold per man or 4% F.C.K. sold per man were considered slightly more refined meas- ures of efficiency of labor than the number of cows per man. The computations involved in the factors consisted of dividing the average number of cows, total milk sold, or total 4% F.C.M. sold by the average number of men. Machinery expense per tillable acre was taken from the Mail-In Accounts where it was routinely assembled by summing the costs reported as repairs on tractors, crop 61 and livestock machinery and equipment, truck upkeep, gas, oil and grease and other machinery supplies, as well as the depreciation on all power crOp and livestock machinery and equipment and dividing by the number of tillable acres. The division by tillable acres was undertaken in an ef- fort to make machinery costs comparable between farms of different sizes. “‘ The percent livestock income from dairying was in- m cluded in an effort to study the existence of a secondary livestock enterprise on farms which were primarily dairy in character. Percent rented land, the last farm manage— ment factor included, was considered an organization fac- tor, and was calculated from information presented in the Hail-In accounts. Acres of rented land were divided by total acres to furnish the factor percent rented land. A total of 58 farm management factors were develOped as a part of this study. hese factors were chosen for use in an effort to compile a list of management factors which would consider all phases of the farm operation. III. Analytical Design and Method The relationship of an independent variable to the dependent variable may be either linear or curvilinear. A linear association is one where a constant amount of in- crease in the dependent variable is associated with a unit 62 increase in the independent variable. Any other relation— ship would be curvilinear. Economic theory precludes either constant returns to scale, constant input—output relationships, and con— sequently constant returns to the inputs of production, or even constant returns across organization or intensity changes over an unlimited range. An assumption of curvi- linearity in some of the farm management factors was there- fore felt to be valid. The association between net income and each of the farm management factors was studied to establish linearity or the degree of curvilinearity. A tabular analysis was performed in which the separate farm management factors were independently ranked from low to high and then divided into 5 equal groups according to the number of observations. The average for the farm management factor and for net in— come was computed for each division. In addition the ob- servations were sorted according to year and studied in the same manner as above. Results of the tabulation in- dicated that one of several basic relationships existed between net income and any given farm management factor and that this relationship was of the following order: Unit increases of the management factor resulted in a constant rate of increase or decrease in net income; unit increases in the farm management factor were accompanied by increases nib 65 in net income at a decreasing rate, a relationship which could be expressed as movement along two or more segments of the curve defined by the law of diminishing returns. The diminishing returns curve may be described as one in which unit increases in the factor result in net income in- creasing at an increasing rate, increasing at a decreasing rate, reaching a maximum, and finally decreasing. A fourth f‘ type was observed which showed little or no relationship between the factor and net income. Tabular analysis then served as the basis for deciding to study the fit of each farm management factor to a linear, second degree, and third degree curve. It was felt that the study of func- tions other than those three would be of no value to this study. Correlations used to study linearity were computed by the use of a Control Data Corporation (CDC) 5600 compu- ter through the implementation of the CORE (10) routine, as were all correlations throughout this study. The basis for deciding which degree of linearity to accept consisted of computing the linear function where Y a + bX, followed by the computation of the function Y = a + bX + 0X2. The 0.. D .p .. 2“ 2 coeifiCients 01 multiple determination, Ry°X and Ry.xx2’ obtained by the two computations were observed. The hypo- thesis was tested that the two coefficients of determina- tion were equal and as a consequence acceptance of the 64 function as being linear. The acceptance of the alterna- . .. . ,_., - “2 tive hypOtheSis that the n values were not equal was the acceptance that the additional variation explained due to the use of the quadratic expression was of significant mag- nitude and that unit increases in the farm management fac- tor caused less than constant increases in net income. , _ . i. -2 T2 a The hypotheSis that R = R , c was tested at the .01 y.x y.xx level of probability. 0 ° 1.1 v ' "'2 r Computation of the function Y = a + bX + CA + dif5 followed and the function was accepted or rejected on the same basis as was movement from the linear function to the quadratic. That is, the hypothesis was tested that 2 2 R ,2 = R, ,fi2 5 at the .01 level of probability. The y.xx y.ix 2 y.xx 3' x acceptance of the alternative hypothesis that R 2 is not equal to R§.xx2x5 is the acceptance of a function which follows the general form set forth by the law of diminish- ing returns. In addition to satisfying the statistical tests, the final functional relationship was expected to approach the expectation arrived at through logical deduc— tion. The prediction model developed was as follows: if. =o(+f(/5.X.. +1{.X..2 +51. .5) + u., where r. = the i 1 1J i 13 i 13 J i ith observation of the dependable variable; Xij = the ith observation on the jth independent variable;cK = a constant term computed from the sample and equal to I - £(fliiij); ON \n fgi = the standard partial regression coefficient of Y on Xi’ which is an expression of the weight given to Xi; X: = the standard partial regression coefficients of Y on ,2 , . , . . a , . t . . -,2 k., wnicn 18 an expression 01 the weight given to A.; i v i Si = the standard partial regression coefficient of Y on XE, which is an expression of the weight given to XE; nd uJ = the random component. The jth variable may be ex- - . 1 r w .2 . .2 .5 ressed as eitr"rvfl32u- .x. + .V.- or .X. + .X. + .X. p i5 1 i’ i i KlAl’ i 1 Ki 1 81 1 dependent upon the relationship of the variable to net in- come. Variation in net income was explained by the function 2 2 2 2 . n , 2 O/y = {R + (u , where {y was the variance 0:: 1, (R was . p . . h . (’2 . the Variance of Y explained oy the regression, and u is u ' ' '1 1 ' ”2 the variance of Y unexplained oy the regreSSion. D y.x was the estimate of O/au and 32y Was the estimate of (372. There- fore, R2 = By2 - 52y.x = l - Sv2.x .... 2 .4; ° b m2 y o y Where variables approach independence, a prediction equation may be most efficiently developed by first includ— ing the entire list of studied variables and then singu- larly removing variables. The equations can then be tested at a predetermined level of probability to determine if there was a significantly greater amount of variation ex— plained when the variable was included. By such a process of removal and test, a prediction equation may be develOped which contains only those variables which offer a definite contribution. The farm management factors used in this study did not lend themselves to such a treatment. A continuum of correlations among farm management factors was to be ex— pected. This continuum could be expected to extend in descending order from those factors which were measures of "‘ the same output, through those which indirectly measured the same characteristic, to those factors which, while they were not directly related, followed similar courses and finally to factors which did approach independence. In arriving at the final prediction equation, the farm management factors were first divided into groups which measured size, crOp efficiency, livestock efficiency, labor efficiency, costs, organization and intensity. The groups were then studied on an individual basis. Factors were studied so as to determine the correlations between farm management factors as well as between the management factors and net income. Classification reduced the number of variables studied at any one time and allowed greater comprehension of the relationships between management fac— tors. Due to the possible relationships betwee groups, more than one set of variables were accepted from groups containing numerous management factors. It was felt that 67 the relative value of a factor within a group might be sub- stantially different from the relative value of the same factor in the presence of management factors from other groups. If such correlations existed, the farm management factors accepted for the final prediction equation might not be those which explained the most variation in net in- come when studied within groups. The sets of farm management factors contributed by the groups were combined in all possible combinations, and the prediction equation accepted was the one which explained the greatest amount of variation in net income. The coeffi- cient of determination (32) was used as the measure of the variation in net income which had been explained, and the equation selected was the one which exhibited the largest coefficient of determination. Farm management factors in— cluded in the prediction equation were removed one at a time and the coefficient of determination was determined using the remaining variables. The hypothesis that the values of R2 for the two equations were equal was tested. Acceptance of the hypothesis was rejection of an increase in the ex- plained variation in net income and the consequent deletion of the management factor from the prediction equation. A study of the effect of breed was carried out. Farm record years were sorted into two breed groups, Holstein and Jersey—Guernsey, and the coefficient of determination was computed using the determined prediction equation. RESULTS AID DISCUSSION The farm management factors which were measures of size had a higher degree of correlation to net income than did any of the other management factors. Correlation co- efficients along with other pertinent characteristics con- cerning the management factors are presented in Tables 4, HI, 5, and 6. Economic theory states that as size increases 2 beyond a certain magnitude diseconomies of scale will re- L. sult in income increasing at a decreasing rate. Decreasing returns to scale were exhibited as cow numbers increased. Acceptance of the second degree curve resulted in pre- dicted incomes for 50, 100, and 200 cow herds which were respectively 5487 above, $725 and $11,877 below that which would be expected under constant returns to scale. It should be pointed out, however, that although the range in cow numbers reported in the study extended to 216 cows, the number of farm record years reporting more than 100 cows was limited, and that the meaning of the regression line at this level should be seriously questioned. The relationships of the remaining measures of size to net income was best explained by linear functions. It is possible that none, or an insignificant percent, of the farms studied were large enough to exhibit diseconomies to scale. It is suggested, however, that the number of till- able acres and number of men are not sufficient in themselves _ 68 - 69 as measures of size, but rather that they can be expected to be of greater value when in combination with other man- agement factors. Total investment is both a measure of size and quality of inputs in that soil and cows of high inherent ability are valued higher than those of lesser ability. This feature of the factor does not lessen its value in a prediction equation, but it does reduce its “" usefulness in a consideration of the effect of size on income. The relationships of crOp efficiency factors to net income are shown in Table 4 and are represented by the variables X6 through X12. The theoretical treatment of an output as represented by crOp yields leads to the expectation that the crop production per acre would fol— low the curve outlined by the law of diminishing returns, and consequently that the total value of crOp would like- wise eventually reach a maximum and decrease. Crop yield index had such an effect on income just as did the sole factor representing a crOp input, fertilizer and lime expense per tillable acre. CrOp value per tillable acre, crop acre value, soil value rating, and percent cash crops exhibited a curvilinear relationship which increased at a decreasing rate when correlated with net income. The relationship between percent tillable acres idle and net income was linear. .omoam e>flpmmoq m nopflQHQXm nopsmaoo floapoq:% one .seflaesssese no seems so. the es esseeeesmsm .N9 + w n N we soammosmoh Hmoqfla How madahow HmHoQom one .mNo + MD + w u M wasshow HmHoQom esp mp domflaepomhmflo mam Mmflhmp Imm one unonpfls HmquHH>HSo mm cepmfla weepomm .mxn + mNo + MD + m u M Show ammosom one mo me Mme Iwopmm one kn dootsnamod whopomm HmDQHHH>Hso one how Gospmflem> pmos esp msamagxo moans soapoqsw mums Q m mso.m seems. hmmmo. m.a m.m eheseg s00\.<.e heessz was eho.m moss. smmo. o.mm m.sa smesfieeesse ages eessem a max sma.m noose. mooo. m.m m.mm esteem sheen seen esoesH .emeq a sea mHH.m mmeo. Hooo. om.sh em.mmw seesflfiseeso .e.a\emeo seesesess max mom.s owmm. sommo. o.ma o.ma *eeeseseeeso meoeo shes a max nmo.m seams. hammo. H.m m.m seesaw mHeH meets massaaee a Has mmo.m emom. tease. om.me om.om .eteseaeesso .a.e\emoo smseaeeeem OHM mom.e mmam. «memo. mme amen seeseas>eso msfleem agate seem ow smm.s moom. someo. so.me mo.mme emeseaeesso heats mess gene we mmu.s mosm. moses. o.Hm m.mm .eheseaeesso xeesH eats» echo he Hmm.¢ mmss. «meme. me.mae mm.mme esessaeesso .4.a\esam> echo ox mmm.s swam. sumem. 5.0 m.H eweseg see he messes me Hmm.s msmm. semmm. so new esteem mesea assesses ex mmw.m mono. shsms. w.mH n.mm eseseaeesso hseo no hesssz ms mmo.se made. emmmm. Hom.osm ems.¢mh esteem esteemeesH Hesse ms mma.mh sua.me msoesH eez HM mwpmgflpmmfl mPQ®HOH.H,H®OO QOHmeflghmpmm WQOHDLNHNer WHOPOMW mamOOQH #QZ WHOPOMW mo Momma noflpmamasoo mo dumpsmpm Ho 0p pnoaommmmé damdQMpm mamapadm mpquOflwwooo manna magquflpmHom seem doflpmmflmmmao cum .pmoo hamsflflomfi .hofimeoawmm moao .mNHm mo moHSmmog .m .memeimmma .mshes eseeem seem asoa .mehsm seems sameness .oSooQH poz op whopomm psesemmqmz seem mo mmflnmmoepmaom i manme .omoam m>flpmmou m 6mg nopsmsoo QOHpoQSw $39 m” .seseeshsoee he seems H0. the eh eseeeeesmemm m00.m posse. mmeo. mse 0mmht sheseaeesso s00\sm00 ewes mmx sm0.s mmsm. «memo. mm ass sm.mmae sheseasesso ewes 00He\esoesH .ehea mmx ma0.m Hm0m. tmmso. mm mes mm.mwe theeeaeeeso s00\esOesH maesso ems 000.: 00mm. smmm0. s.s m.ms semseaeeeso sass them a 0mm 0ma.m 0Hs0. ma00. 0m 0 m0.0 theses ewesm sooneseseeeseem 0mm 50H.m omo. mmoo. meme mmsm shessasees0 soo\mesese>oseeH mma me0.s them. emsee. Hm.0e he.se seeses es0\eewhe see we ems 0H0.m whom. soms0. ms.0e m0.es hasssseeeso sso\eeesm sass ems mom.s m0mm. semm0. em owe esteem A.meav s00\psm seas mma sm0.e e0nm. ammuo. ems.a 0m0.0a whenqu.msav s00\eeeseosm 20m as ems m w.s smam. mmsoa. has.a mmm.0a eweseq A.msav s00\eaom 20m me mm» mm0.s whom. sso00. 003.H msm.aa theses A.msav s00\eeeseeem sees mmx 0mm.s weom. wHH00. mme.a oe0.0H seesaw A.mssv soo\eeem sees ems 005.: smmm. enema. mew mmse esteem s00\meahm sheen 0mm mms.e ¢s0m. smsmm. mmm,m0 www.mom sateen n.meav ses\eaom nos n max mes.s mama. swasm. nm0.Hn mmw.oam hashes A.msav ses\eaem sass max ems.s 50mm. moses. s.m m.0m theses ses\meos senses use Gasmaflnwmfl mPQmHOHMMOOO QOHPMQHEumwQ mCOH#6H\V®Q mrHOPomm mamOOQH me mHOfiowpm mo Hosea Goapmaosmoo mo damwsmpm go on pamaommqwz nhmommpn mamflpH33 memofloflmwooo mused madeoHpmaem Sank m eases .thOHonma Moopmo>aq one honofloflwwm Honda mo whammmoz .n .mmmaummma .meses eeeeem sham Hs0a .mssse seems semeseez .osoqu pom op mwopomm psosommsmu swam mo mmflmquepmHom 72 .omoam o>flpmmoq m 6mm dopsmsoo floapomdw wee .sesaesseoee so Hosea H0. the em esseaeesmane .N9 + m N we Qoflmmmsmom hmomfla fi How wasasoy ammonom one .uxo + N9 + m u M masswom ammosow own an doNHHopoMHmno ohm Mmflmepmw one piompfls hmquHH>MSo mm nopmfla maopomm .mNp + two + N9 + m u M Show Hmmmsom esp mo ma Mmfl Isopmm one 59 nopdnwflmon whopomw Hmmfleaa>wso map How moammflhm> pmos map mqflmamxo flown; soapossm one* mam.m 0m0H. smmmo. s.m 0.mm useseaeesso “.msav se0\eoo\zme 0mg . A.mfl.mv 4 mm sum m ones. mma0. m.m 5.0m seeseaseeso sh0\s00\eseae>HsOm hem M mmm.m when. mesme. ham mes.m eatees A.mssv eoe\semee ems mna.m smmmm. smsmo. m.mm m.HmH .seessaeeeso essemem so when 0.x enm.m 0haa. sasao. u.m s.om seesaw sass he hseo a mmx mmm.m mama. sommo. mm.ome Hm.Hth hmeseg emoo ewes emseees emx opmmflpmfl mpsoHOflwmooo QQflpmsstopem mQOflDMHbeQ mmopomm oaooQH pom mmopomm mo poshfi moflpmaoasoo wo nemeswpm Mo op pdmsomwsmfi pmmdsmpn mamapada mevHOflwwmoo mflmmé QHSmQOHpmaem 89mm .Mafla QH msoo psoohem new mpquH doom mo mmMSmmoE .o .Nwmalmmma .mwmow whoomm 89mm dam .mshmm hsflmm qmwflfloeq .oSooQH pom op meopomm psosommqma 89mm mo mmHQmQOHpmHom 0 oHQmB 75 Cropping factors failed to have the degree of cor- relation or to decrease the standard error of estimate to the extent exhibited by size factors. The contribution of several of these was meaningful, however, and indicated a definite relationship between crOpping system and income. As shown in Table 5, all management factors measur— ing the production of dairy products on a per cow basis as well as the dollar returns for the sale of these products were linear. Any decrease in the quantity of milk produced per unit of input with increasing levels of production was not reflected in the correlation of dairy output factors to net income. The relationship between the price of milk and net income was such that net income increased at a decreas- ing rate as price increased. This curvilinearity can be contributed to the relationship between fat test and price, which in turn can be substantiated by the linear relation- ship between price and net income when price was calculated on a constant fat basis. The aesociation between the high fat test and the corresponding lower level of milk produc- tion was further borne out by a tabulation according to breed. A total of 909 farm record years for Holstein herds had an average fat test of 5.66%, produced 11,769 pounds of milk per cow, and received an average net income of 38,408. The total for 152 farm record years for Jersey herds showed an average test of 5.00% with 8,574 pounds of milk per cow 74 and was accompanied by a net income of $6,562. This dif- ference of $l,870 in net income was responsible for the greater part of the curvilinearity observed when milk price was correlated with net income. The livestock efficiency factors which were meas- ures of feed consumption are listed on Tables 5 and 6. Feed cost per cow was a factor calculated from farm account records and the coefficient of determination of .0152 indi- cates that the factor had no value in this study. he method of computation accounted for the meaningless re— sults. Average feed cost, grain, hay, and TDN per cow were all measures of feed input. These measures all had low correlation coefficients, as was to be expected, as in reality they were secondary factors——secondary in the sense that they were factors which influenced milk produc- tion which in turn influenced net income. The farm management factors measuring labor effi— ciency were second only to size factors in their degree of correlation to net income. The higher correlation for the two measures of milk sold per man over cows per man was not unexpected and reflected the effect of level of production as well as the efficiency involved in handling cow numbers. Those management factors designed to measure organi— zation or intensity showed surprisingly little correlation to net income. Improvements per cow, replacement to cow 75 ratio, machinery cost per tillable acre, percent livestock income from dairy and percent rented land all lacked sig- nificance at (P < 0.01). The farm management factors total investment, number of cow, number of tillable acres, and number of men were all measures of size of the farm operation. It can gen- erally be stated that increased herd size demanded in- creased acreage to furnish a larger quantity of feed and an increased number of men to handle the larger herd and farm. As number of cows and/or acres increased total in- vestment was also required to increase since the two fac- tors, cows and acres, were a part of the computation of in- vestment. The coefficients of determination computed when two or more of the measures of size were jointly correlated with net income demonstrated, however, that the relation- ship was not constant. The addition of total investment to number of cows increased 32 from .4547 to .4580 while the addition of tillable acres to number of cows increased R2 to .4445. when tested to determine if the added vari- able explained a greater amount of variation both total in— vestment and tillable acres were significant at the .01 level of probability. The number of men did not signifi- cantly (P <.01) increase R2 over that for number of cows alone, nor was the addition of number of tillable acres to a function utilizing number of cows and total investment significant (P < .01). 76 The functions accepted as measures of size were as follows: Y = f(X2,X5,X§) and Y = f(X5,X§,X4), where Y = net income; X9 = total investment; Xan = number of cows; and X4 = number of tillable acres. The combination of cows and acres had a lower coefficient of determination than did the combination of cows and investment, however, it was retained to study its value when combined with fac- tors representing other characteristics of the farm opera— tion. C Crop value per tillable acre was arrived at through the multiplication of price and crOp output, followed with division by the total tillable acres. Consequently, the factor was primarily a product of yields and values with consideration indirectly being given to the degree of utili- ,zation of land. The aspect of land utilization entered. into crop value per tillable acre as a result of idle land being included as a part of total tillable acres. It fol- lowed that crop value per tillable acre might be considered in the role of a primary factor in the explanation of Variation in net income as a result of cropping practices and crop yield index, crop acre value and percent tillable acres idle serving as secondary factors. It was felt that if crop value per tillable acre was used in predicting net income that the three secondary factors could not logically be included. This logic was extended to include soil value 77 rating and fertilizer and line eXpense per tillable acre which were considered to be tertiary factors in the ex- planation of variation in net income and primary in the explanation of variation in crOp yield index. The concept, as stated, was tested by correlating the secondary and tertiary factors with crop value per tillable acre in such a manner as to receive a coefficient of determination after the addition of each factor. The function studied was as follows: X6 = f(X7,X 2X5 X8 ,Xg ,X5 X -X X g), 7’ 7’ ’ ll’ 10’ 9’ where X6 = crop value per tillable acre; X7, X7, X3 = crOp yield index; Xg’X§,Xg = crop acre value; Xll = percent tillable acres idle; X10 = fertilizer and lime expense per tillable acre; and X9,X§ = soil value rating. Prior to studying the above relationships the degree of curvi- linearity was determined for the factors when singularly correlated with crOp value per tillable acre. The curvi— linearity established was that shown in the preceeding func— tion, where both crOp yield index and crop acre value were best explained by a third order curve; soil value rating by the quadratic; and percent tillable acres idle and fer— tilizer expense with a linear function. The coefficient of determination was .6264 when crop value per tillable acre was correlated with crOp yield in— dex. with the addition of crOp acre value, percent till- able acres idle, fertilizer and lime expense per tillable 78 acre, and soil value rating the coefficient was raised to .8808, .9251, .9258, and .9285, respectively. Each of the five factors was significant at the .01 level of probabil- ity. The high percentage of explained variation in crOp value per tillable acre bears out the concept of crOp yield index, crop acre value and percent tillable acres idle as secondary factors. The acceptance of this concept logically leads to the examination of net income with either crOp value per tillable acre or a combination of crOp yield index, crOp acre value, and percent tillable acres idle. An attempt to increase the variation explained in net income by adding the secondary crOp variables con- firmed this, as none of the three were able to signifi— cantly (P < .01) raise the coefficient of determination. Fertilizer and lime xpense per tillable acre was of no value in increasing R2 when added to either the primary crop factor or to the secondary factors. Soil value rat- ing failed to follow the format set down for secondary or tertiary crOp factors and was of significant (P < .01) value in explaining variation in net income when added to either crOp value per tillable acre or to the combina- tion of secondary factors measuring yield, value, and land utilization. The value of the factor extended beyond that of a simple measure of soil productivity. 79 The following measures of crOp efficiency were ac- -_ 2 2 2 cepted for future study: = f, Y = faompo mo Hmpadq was .mmm u mama momm mnmm u coma mmom u mmma "mo H 2 . . . . . ea mm m m m m e m a m o.e m.m 3.0\.a.a 5.9asz a am.na ao.ma ae.ma ae.ea am.ea ae.ea eemg empatm a max new new new mum ago aum anamm ea Scam meoqu .pm>q & N am.mma Ao.ema we.mma oa.mma ma.ame Ho.mme .4.a\pmoo amquAowa max am.ma am.ma ao.mH ao.ma ao.ea au.ma macho amwo pqmowmm max a am am an ea am mHeH mowoa manmfiaae a Ham em.ew no.em me.ea mm.ma em.mm mm.mw .4.a\pmoo pmuwaapwmm oax Hmam Hmaa mafia cmaa Hmam mmaa meapmm msam> Haom ma 5mm Sew Hoe Hma emm mma msaa> whoa mono ma mm mm mm mm mm mm AmeqH eaoaw mono LN ama ama mam men mma Hma .<.a\.sH.> mono ex o.a o.m o.m o.m m.H m.H awe no ampasz ma mam omm mam mom mom ema menoa magmaaae ex u.mm a.me e.oe e.em o.mm m.mm .300 no Ampasz ma em¢.¢ma mee.ama Leo.emm mam.mme Hmm.nnm Hmm.mea meApmm>QH Hmpoe mx eua.mm emm.ma mma,mm www.ma. moo.ma emm.na meoqu pmz Ha ammaasm mead Hema coma mmma mama whopomm new» anew pamammmqms e>flm H Shem .QOHpmmflqmmao Sam .mpmoo haemamoma .theHOmem mono .eNHm mo mehdmwmg .m .mamow Shoomm 89mm adoa .msamm hpflmm ammfiQOHE whopowm panamwmnmz 89mm How medam> makeow mwmhm>¢ H magma xaeqeaaa 113 .mmm u coma ”mom mmma n mmma “maoaaom mm mam 9mm» mp mQOflpmbnmeo Ho Hmnafiq mfla H momm mama momm goo\pmoo @mmm mmx mmfim mmam mmam womm ooam\maoqu Moopmm>flg mmx mum mmm mmm goo\maooaH mappmo me mm.am Rm.H¢ mm.m¢ Mafia mmmm & omx mm.o mm.o mw.o OHpmm aooupq®8momammm mmx mmmm mmmm mmmm 300\mPQmam>oumaH wmx om.¢m mo.¢m mo.¢m ezo\m0Hum 20m mm mmx oa.¢m Hm.mm mm.mm aao\m0Hnm Mafia mmx .mpa mm: .mpa mm: .mpa mm: aooxpmm xafla mmx .mnH omm.oH .mpa Hom.0H .mpH omm.oH aoo\omosmoam gum §¢.¢mx .mpa mom.oa .mpH Hmm.oa .mpa Hmo.oa aoo\uaom 20m mm mmx .mpa mHH.HH .mpa mom.aa .mpa mom.HH aooxmmosuopm xHflz mmx .mpH mom.oa .mnH mam.oa .mpH mmm.oa aoo\caom Mafia me mmmm Hamm momm aoo\mmfimm mgflmm omx .mpH mma.oom .mpa mmm.mma .mpfl mmm.mmH gmzxufiom 20m m: max .mpH moa.mom .mna mmm.mom .mpa mmm.¢mH amz\oaom Mafia max m.ma o.mH m.ma qm2\mgoo “@9552 max coma mmma mmma muopomm Hmmw pamammmqwz H 89mm hommHUflwwm Moopmmbflg dam hommHOHmwm 909mg mo mdemmmz .m m magma xflwnmmm¢ madmw whoowm 89mm H¢oa .mamwm huflmn qmmfl£0H§ mmopoww pamammmqma Shwm mom mmdam> hahmmw mmmpm>¢ 114 .mmm n mwma mmmm n Hood “macaaom mm mHm Hmmh hp mQOHpm>Hmmpo Ho Ampadn age a ommm mmmm mmmm aoo\pmoo cmmm mmx mmam Hmfim mmam @mmm ooam\maoqu Moopmm>fiq mmx mmm mmm Hmm g0330qu mappmo me mm.m¢ Rm.am mm.m¢ Mafia mmmm m cmx mm.o mm.o mm.o OHPmm gooumeamomHmmm mmx ommm mamm momm aoo\memam>onmaH mmx ma.¢m HH.¢m mm.¢m aao\mOHHm aomnmm_mmx mo.¢m mm mm om.¢m emo\00flgm xaflz mmx .mpa mmm .mnH mm: .mna mm: goo\pmm xHfia mmx .mna omm.oa .mpa mom.HH .mpa mom.oa aoo\umoscogm 20m & mmx .mpa mmm.oH .mpa mom.oa .mna mom.oa aooxoaom 20m mm mmx .mpa m¢m.aa .mnH mmm.aa .mpa omm.HH 300\cmoswogm MHHH mmx .mpa mmm.oa .mna mmo.HH .mpa omm.oa aoo\waow xHflE me mmmm mmmm mmmm aoo\mmamm mnflwm omx .mpa mmm.mom .mpa mmm.mmm .mna mmm.aam qm2\oaom 20m m: max .mnH www.mam .mpa omm.¢mm .mpa mom.mam qm2\caom Mafia max m.om o.mm m.om qm2\msoo umpasz mam hhmaadm mama mea whopowm Hme Hmmw pumammmqmz m>Hm H 89mm deQHpQOOIIm magma Nflcnmmm< 115 .moH u mmma mmma u amma mmma u omma “mmfl n mmmfl mwm u wm®H umBOHHOH mm. me .Hmmkn hp mQOHPm>HmmDO .HO .Hmflgfl 0H3... H .mpa o.mm .mpa m.mm .mpa m.mm mmm\aoo\zme mmx .mpa am .mpa om .mpa am mmm\aoo\psmam>asmm mmm mmx .mpa m¢¢.m .mna mmm.m .mpa mmm.m goo\Qamao mmx mma maa mma manpmmm no mama mmx mm.mm mm.mm mm.mm aaaz ma maoo & mmx mmam amam mmam pmoo @mmm mmmam>¢ mmx mamaesm mama amma mew nmmw m>HwW H .mpa m.mm .mpa m.mm .mpa m.mm me\goo\zne mmx .mna am .mpa am .mna am mmm\aoo\meam>asmm mmm mmw .mpa m¢m.m .mpa maa.m .mna moa.m 300\qamao mmx mma ama mma manpmmm no mama mmx Rm.mm mm.mm & .mm aaaz qa meoo & mmx mmam moam aoam pmoo ommm mmmao>a mmx coma mmma mmma myopomm Hmmw pamammmnwz H ahmm MHHS SH msoo pamohmm flaw hoquoHme dwmh mo moHSmmmE .m m magma xaGQmmma mammw whoomm 89mm #Hw .maamm mHHmm nmmHQOHE whopomm pmmamwmnma Sham How mmsz> hHHMmM mmm9m>< 116 .mmmmpow prop H0 &H qmnp mmmH u. .maom H qmgp mmmH * xxx xxx mam oNam anmm mwmnm>< mmoao Hm>oo .mmmnpv II ** * mdomquHmomHa II ** * mmHHHmm dam mpHsam II ** H macho M0599 I: m t Mnmm HHom II m u BOHHmm Ho meH Ammopmpom .mpmmp HmMSm .mqmmnv u: m m mmoao gmmo amgpo .3 mm m ma p881 Amha .mpHmmm .hmHHva II H m qumaw menmo Hmnpo .59 00 OH Hm mpmo .:9 mm ma mm gamma qaoo mace m.aa aa mm mmmaam qaoo mace m.m NH om maSpmmm mHQMHHHB Ammg doNHa no mMHmMHwV mace o.m am we mmmnom onod 9mm memo qH 89mm @090 QH mmao< Sacha mayo UHmHM mo mo mmmnm>4 ammonmm Hmnadz mmmnm>4 .momlemmH .mHMmM whoomm ahmm H¢OH .manmm mnHmm nmeQOHE qupmhmmooo no macaw mmoao mo memHM dam .mmmMPQmonmm .mmmwm904 mmmnm>¢ m magma mammmmma 117 ‘Jfllw .mmm u omma “wow u mmmH ”mm H mmmH “mBOHHom mm mam mem hp quHpm>ammpo Ho Hmnadm 059 H oo.mm mm. oo.om om. oo.mm mm. mpammm oo.ma mm. oo.¢a oo.a In- I- mam oo.mm ma.a oo.m oo.a sun I. pmmgaaosm oo.mam ¢¢.a oo.ma¢ oo.a 00.9mm mm. mmopmpom oo.mma oo.¢a oo.mma oo.aa oo.m¢a om.aa mammm ammsm oo.om mm.a oo.w¢ oo.m oo.¢¢ mm.a mammm mom oo.mm wa.m oo.¢m mm.m oo.mo mm.m mammm oo.am mm.a oo.m© mm.a oo.mm om.a pawn: oo.mm oo.m oo.am oo.m oo.ma oo.m manpmwm mapmaaae oo.mm oo.ma oo.m¢ o¢.ma oo.a¢ oo.ma mam Umxaa oo.om ow.wa oo.mm om.om oo.mm oo.om mam maamaaa oo.mm oo.m oo.m¢ oo.m oo.am oo.m mmmaam mmmam oo.mm mm. oo.¢m om. oo.a¢ om. mmaamm oo.mm mm. oo.mm om. oo.wm om. mpmo oo.mm mm. oo.mo oo.a oo.mm oo.a mamam qmoo oo.mm m oo.m m oo.om m oo.m m oo.om m oo.m m mmmaam qaoo msHm> pHGD\ mSHm> pHQD\ msHm> pHQD\ QSOHw mHo< pQMpmaoo mpod pampmuoo mHo< pddpmmoo QOHU momo oOHHm @090 mOHHm mono mOHHm omma mwma mmma Hawmw .mGOprbnmmpo choomm 89mm H¢OH .manmm hHHmm Qme£0HS qupmammooo How msHm> oao¢ mono cam @904 mHQmHHHB Hmm mSHm> @090 quBHmpmm 0p ©mmb mw mdew> dam mmOHHm @090 m magma xaGQmmm< 118 .mmm u m0ma mmmm u a0ma “mgOHHom 00 090 900% mp mQOHp0>HmeO Ho Hmnadq was a 00.00 mm. 00.00 mm. 00a000 00.mm 00.a 00.00 00.a 000 00.00 mm.a In: up 000030000 00.000 0a.a 00.00: 0m. 00000000 00.0ma m¢.ma 00.0ma mm.ma 00000 00000 00.m¢ 00.0 00.00 00.0 00000 000 00.00 00.0 00.0m 00.0 00000 00.00 0m.a 00.00 mm.a 00002 00.ma 00.0a 00.00 00.0a 0000000 mapmaaae an- In: In: nu: 000 omxaa 00.00 00.00 00.00 00.00 000 00a00a0 III II III II 0m0HHm 00090 00.am 00. 00.00 00. R3200 00.00 00. 00.00 00. 0000 00.00 0m. 00.00 0m. mamm0 0000 00.0m 0 00.0 0 00.00a0 00.0 0 00maa0 quoo 03H0> pHmD\ 0SH0> pHflb\ macho 0904 pampmnoo 0904 pcwpquo @090 @090 moHHm @090 mOHBM m0ma a0ma Hummw dmsQHpqOollm 0H908 NH0Q0mm4 Appendix Table 6 119 Range in Values Observed for Farm Management Factors. Cooperating Michigan Dairy Farms, 1041 Farm Record Years, 1958-1962. Farm Lowest Highest Management Value Value Factors Observed Observed Xl Net Income -$8,485 $59,108 X2 Total Investment $9,927 $324,926 X5 Number of Cows 9.9 215.7 X4 Tillable Acres 60 666 X5 Number of Men 0.8 6.0 X6 Crop Value/T.A. $15 $121 X7 Crop Yield Index 52 165 X8 CrOp Acre Value $51 $98 X9 Soil Value Rating $35 $500 X10 Fertilizer Cost/T.A. $0.00 $20.96 Xll % Tillable Acres Idle 0% 49% X12 % Cash Crops 0.0% 61.9% X13 Machinery Cost/T.A. $5.07 $55.85 X14 % Lvst. Income from Dairy 71% 100% X15 % Rented Land 0.0% 100.0% X16 Number of T.A./COW 1.1 15.} X1’? Number Cows/Man 7.0 47.9 X18 Milk Sold/Man 70,958 lbs. 512,852 lbs. X 4% FCM Sold/Man 49,407 lbs. 490,545 lbs. I—‘ \O Appendix Table 7 Range in Values Observed for Farm Management Factors COOperating Michigan Dairy Farms, 1041 Farm Record Years, 1958—1962. _—— 120 1 :— Farm Lowest Highest Management Value Value Factors Observed Observed X20 Dairy Sales/Cow $212 $611 X21 Milk Sold/Cow 5,596 lbs. 15,179 lbs. X22 Milk Produced/Cow 5,555 lbs. 16,174 lbs. X25 4% FCH Sold/Cow 5,289 lbs. 14,556 lbs. X24 4% FCM Produced/Cow 5,158 lbs. 15,594 lbs. X25 iilk Fat/Cow 195 lbs. 595 lbs. X26 Milk Price/CAT $2.45 $5.66 X27 4% FCM Price/CWT $2.58 $5.05 X28 Improvements/Cow 344 $1,424 X29 ReplacementzCow Ratio 0.0 2.75 X50 % Base Milk 22.4% 55.5% X31 Cattle Income/Cow -$24 $406 X52 Lvst. Income/0 Feed $81 $575 X55 Feed Cost/Cow $45 $504 X54 Average Feed Cost $111 $290 X55 Percent Cows in Milk 74.1% 95.5% X56 Days on Pasture 0 217 X5,7 Grain/COW 1,000 lbs. 7,455 lbs. X58 Hay Equivalent/Cow/Day 14 lbs. 55 lbs. X59 TDN/Cow/Day 15.5 lbs. 54.4 lbs. I For farm management factors with D.H.I.A. source of data, the farm record years are only as the limited to 814. 1’ 1" MUM USE ONLY ‘0 '1'. Vw'. ‘9 my. a..»",l . x .' , , - a . H“: ' t .N', fie ‘ V .43 J 113.. ._‘, . u;\§ A. $104 ' "1111111111!“11115