f1: 9*, ..1: z “u. '4 .1 . “ star :55; ' ES" . 1"; >334 1' it: 1 . 1w .3. W 1. ..M . 2 N . mm mm ban? Vuw VTA Li‘hfl. 1 . ANAWWVI Id 4d.» 1.1%th .. xiv? .m‘mmafi , w...14?uww.muws .94.. yiww‘ 1W1, , .{umr 39mph “sun . \uwm‘flvmwn. m ,rJn .anbfih‘w «[541! 1‘“? PM.» « ll‘llllllllllll ”23's LIBRARY Michigan State University This is to certify that the thesis entitled A COMPARISON OF EXPENSES ON BOARDING, RIDING, AND TRAINING STABLES IN MICHIGAN presented by Colleen M. Brady has been accepted towards fulfillment of the requirements for Master of Science degmmin Animal Science Major professor Date 2/2/95 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove thie checkout from your record. TO AVOID FINES return on or before dete due. 052 DATE DUE DATE DUE DATE MSU Ie An Atflnnettve Adlai/Emu Opportunity lnetttuton mm: A COMPARISON OF EXPENSES ON BOARDING, RIDING, AND TRAINING STABLES IN MICHIGAN BY Colleen M. Brady A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Animal Science 1995 ABSTRACT A COMPARISON OF EXPENSES ON BOARDING, RIDING, AND TRAINING STABLES IN MICHIGAN BY Colleen M. Brady Data collected by the Michigan Equine Monitoring System were used to analyze the expenses reported on boarding, riding, and training stables in Michigan, in 1990. These data were analyzed to detect differences in expenses by operation type, size category, and size category within operation type. Statistical Analysis Software was used to perform this analysis. The data indicate that there are more differences present in expenses based on operation type than on size category. In most expenses, there was the greatest annual expense per horse on training stables, and.the least on boarding stables. 'There are more differences by size category within operation type, than by size category alone. 0n training stables, there is evidence of economy of scale, while boarding stables show the facilities of 3-9 horses to have the lowest annual expense. Riding stables showed very little difference based on size category. ACKNOWLEDGEMENTS If I singled out every person who helped me with this project, the acknowledgments would be longer than the thesis. First I must thank my committee members Dr. Joseph Rock and Dr. Andrew Skidmore for taking the time to help‘with this project. A special thanks goes to Dr. James Lloyd for going above and beyond the call of duty in helping me understand.the MEMS data, and, teaching' me hOW’ to 'use the statistical software. The debt I owe to Dr. John Shelle I will never be able to repay. His guidance and advice through out this project, and my undergraduate career, have been invaluable. Without his support and "listening ear" I never would have made it this far. I would also like to thank everyone in 124 Anthony for their assistance over the last two years. Especially Dr. Dennis Banks for his statistical assistance, Kim Dobson for her help with computer generated problems, and Camie Heleski for her perspective. I would also like to extend a blanket thank you to all iii the other graduate students who have been there both in class and outside of it. The moral support is invaluable. A final thank. you. goes to 'my friends and family, especially’ my' mother and. grandmother, for all of their support. I would not have accomplished what I have without them. iv TABLE OF CONTENTS Introduction...............................................1 Literature Review ....................................... ...9 materialsandMethOdSOOO0.000......0....00.0.0000000000000018 ResultSOOOOOOOOOO...O0.00...OOOOOOOOOOOOOOOOO0.0.0.0000000026 DifferencesbyOperationType.........................35 DifferencesbySizeCategory..........................40 Interactions between Operation Type and Size Category.41 Discussion..... .................... ........................50 The Effect of Operation Type on Selected Expenses.....54 The Effect of Size Category on Selected Expenses......67 The Effect of the Interaction of Size Category and Operation Type on Selected Expenses.............60 Summary and Conclusions. ......... ....... . .......... .....90 Recommendations............................................92 Appendix..... ........................................ ......96 Bibliography .................................. .............97 LIST OF TABLES Table 2.1 Number of Original Registrations from 1971-1993 ........ 15 Table 3.1 Expenses Selected for Analysis ................. . . . . . . . . 22 Table 4.1 Distribution of Operations by Operation Size ....... . . . . 27 Table 4.2 Income and Expenses by Operation Type. . . . . . . . . . . . . . . . . .28 Table 4.3 Total Dollars Spent by Operators of Michigan Horse operationSe .......... OCOOO0.0.0.0.....00000000000029 Table 4.4 Distribution of Operations in this Study ........... . . . . 30 Table 4.5 P Values of Selected Expenses after Analysis of varianceOOOOOIOOOOOOOOOOOI0.00.00.00.000.0.0.0000031 Tables 4.6A-C Table 4.6A Ranking of Expenses on Boarding Stables by Adjusted-LogMeans................................32 Table 4.68 Ranking of Expenses on Riding Stables by Adjusted-LogMeans....... ..... 33 Table 4.6C Ranking of Expenses on Training Stables by Adjusted-LogMeans ...... 34 Table 4.7 Actual and Adjusted-Log Means of Selected Expenses Compared by Operation Type ..... ...................35 Tables 4.8A-C Table 4.8A Ranking of Expenses on Small Operations by Adjusted-LogMeans................................3‘7 Table 4.88 Ranking of Expenses on Medium Operations by Adjusted-LogMeans ...... 38 Table 4.8C Ranking of Expenses on Large Operations by Adjusted-LogMeans................................39 vi Table 4.9 Actual and Adjusted-Log Means of Selected Expenses Compared by Size Category.........................40 Table 4.10 Actual and Adjusted-Log Means of Selected Expenses Compared by Size Category within Boarding Stables ..........................................42 Table 4.11 Actual and Adjusted-Log Means of Selected Expenses Compared by Size Category within Riding Operations ................ ........................43 Table 4.12 Actual and Adjusted-Log Means of Selected Expenses Compared by Size Category within Training Operations........................................44 Table 4.13 Actual and Adjusted-Log Means of Selected Expenses Compared by Operation Type on Small Operations. . . . 46 Table 4.14 Actual and Adjusted-Log Means of Selected Expenses Compared by Operation Type on Medium Operations. . . 47 Table 4.15 Actual and Adjusted-Log Means of Selected Expenses Compared by Operation Type on Large Operations. . . . 49 vii CHAPTER I INTRODUCTION The relationship between horses and humans is truly unique. For centuries, the horse provided transportation and power to conquer the land and enemies. The introduction of the horse to warfare gave men like Alexander the Great the ability to conquer new worlds, and gave people living miles apart the ability to exchange ideas and information. Not only were horses invaluable in war, they also enabled farmers to better produce crops, and provided entertainment to rich and poor alike. Horse racing, polo, and bashkari (Afghanistan) are but three of the hundreds of games men and horses play together. With the advent of the motorized vehicle at the turn of the century, it appeared that horses would no longer serve as a significant part of the economy. With tractors to replace them in the fields, cars to replace them on the highways, and tanks to replace them on the battlefields, horses seemed destined to go by the wayside, and the sight of the horse would become as rare as the bison. There was one place, however, where the horse was not replaced by mechanization: pleasure. "Paradoxical as it may seem, in our automated, computer-governed age, the horse is once again becoming an important part of American life" (Alampi,1970). As the world became more mechanized, leisure hours increased. This time needed to be filled, and many 1 2 people used.horses to do so. Some bought horses to race, some to show, and even a few to breed more horses. The largest portion of the horse owning population, however, owns horses just for pleasure. To go for a ride on a sunny afternoon, or to trail ride with friends. These people are the heart and soul of the horse industry, and contribute greatly to the economies where they live. Instead of decreasing in numbers, the horse increased in population after World War II, with an estimated 7-10 million horses in the US in 1970. In 1970, it was estimated that a total of $7.5 billion was invested in the US horse industry (Bolt, 1970). It was estimated that the value of horses accounted for over $2 billion, and the remaining $5 billion was spent in the economy to maintain and support the horses. This money was spent at boarding stables, tack and feed stores, and other horse related enterprises. A survey in 1984, conducted by ‘the .American. Horse Council, reported that 750,000 Americans spent approximately $15 billion on the horse industry. This money was spent maintaining the estimated 5.2 million horses in this country, and by spectators attending or betting on horse events. The $15 billion earned by the horse industry in 1984 constituted 16% of the Gross National Product of the agriculture and fisheries section of the economy (American Horse Council, 1987) . In addition to dollars spent by spectators and owners, the horse industry makes contributions to the economy 3 through jobs, ranging from grooms and trainers, to construction workers to build horse facilities, and maintenance crews for Showgrounds and racetracks. There are also people employed. working for tack and feed stores, veterinarians, and publications which interact with the horse industry. The horse industry in Michigan contributes over 256 million dollars to the Michigan economy each year. These dollars are generated by the 28,700 owners and operators that own or care for the state's 130,000 horses. These owners and operators also realize $122,000,000 of income yearLy. The largest portion of this income, 34%, is generated by 1600 riding, training and boarding operations, with a total of 42 million dollars earned.(MEMS 1991) This is consistent with a 1971 Michigan study which also reported boarding as the primary income producing facet of the horse industry. The horse industry is unique among livestock enterprises, in that the most profitable operations are generally not animal production facilities. Although there are many owners/operators who breed horses, and sell them, the most consistently profitable industries are service oriented, such as.boarding, training, or riding facilities. (Holleran, 1973) There are many contributing factors to explain why production facilities are generally less profitable than service facilities. One is the longer generation interval for horses. Where a dairy cow calves and begins contributing to 4 the operation's revenue at 24 months, it is extremely common for mares not to foal before the age of four. This means the operator has an additional two years of input costs before a mare can provide any income. In addition, the foal that mare had at the age of four will likely'maximize it's marketability after being started under saddle, usually between the ages of three and four. In the best case scenario, where the foal is sold as a weanling, the mare is five before she creates a salable product. It is feasible,however, that the mare could be seven years old before she provides any revenue for the operator, if the sale of the offspring is delayed until it is trained to ride. The input costs for producing foals is also higher than for other species. The breeding fee for an outstanding dairy or beef bull is unlikely to exceed $250, while the stud fees for stallions can range from $250 to over $100,000 for an established sire of racehorses. Examination of advertised service fees in several breed publications shows the 1994 average to be between $1,000 and $2,000 for a stallion with a respectable show record. In addition, there are many rules and regulations determined by different breed registries and organizations limiting the use of technologies such as artificial insemination and embryo transfer which are common in other livestock species, and contribute to their profitability from a production standpoint. Turnover of animals in the horse industry also occurs at a slower rate 5 than in other livestock species. Most horses stay in the industry until they die a natural death, which can be in excess of 20 years. Therefore, the replacement rate for horses is much lower than in other species. For these reasons, traditional evaluation methods may not accurately characterize the horse industry. For most horse owners, the horse is a leisure or recreational investment, much like a boat or a summer cottage. The horse's major purpose for these people is to provide relaxation or entertainment after a day's work (Michigan Equine Survey, 1972 & 1984). Because many of these people choose not to be intimately involved in the care of the horse, there is a demand for knowledgeable, competent people to operate facilities to care for or train these horses. It is in this service capacity where the horse industry becomes profitable, and provides a livelihood for its participants. Boarding, riding and training stables are operated with the intention of making a profit, or at least breaking even, depending upon the mission of the operation. For these reasons, this research will concentrate on boarding, riding, and training operations. Three hypotheses will be tested in this study: * a difference exists in expenses based upon operation type, regardless of operation size * a difference exists in expenses based upon operation size, regardless of operation type * a difference exists in expenses by operation size, which varies by operation type 6 This study is concentrating on analysis of expenses instead of income for two reasons: 1) Profitability of a business is most affected by input costs. (Harsh, Connor & Schwab) Profitability is defined as the difference between revenue and costs. Because the facilities considered here are service instead of commodity oriented, the operator has a greater degree of flexibility than does a producer of other animal products, which are sold for a competitive price. Efficient management of costs is still imperative, as it *will increase the profitability of the operation. 2) The expenses reported in the MEMS Survey, which constitutes the data base for this analysis, would appear to be more reliable than the incomes. The barter and trade system are still used commonly in the horse industry, and the value of these items is likely to be unreported, or reported inaccurately. Many income producing activities, such as lessons, are remunerated in this way. In addition,there was a greater degree of non-reporting on the survey questionnaire in the area of income than in expenses. There are some drawbacks intrinsic in the use of survey data, such as reliability of reporting; 'There are also likely to be some inaccuracies in areas such as labor hours reported. As in other livestock species, operators tend to discount their own hours of labor and report only hired labor hours. 7 Considering the paucity of data presently available regarding horse facilities, these analyses will be of value as a starting point for future studies. Surveys or studies quantifying the economic impact of the horse industry are scarce, and analyses of management practices are non—existent. Most surveys concentrate more on how the horse industry affects the economy of an area, than the actual economics of owning or managing a horse operation (AHC,1987, & Economic Impact of the Virginia Horse Center, 1991). There have been studies conducted within other species (Hoyt,1955; Pickler,1947), that attempt to quantify expenses, and make management recommendations. The sheep industry and beef industry both have handbooks designed to help operators evaluate the economic and financial condition of their operations. The dairy industry is by far the leader among large farm animal species groups in economic analysis, and the operators in this field have several references available both on a national and regional basis. Telfarm is an example of a program which assists the producer in analyzing the financial condition of the operation, and utilizing that information to make management decisions. In a search for journal articles discussing the economics of various livestock enterprises, there were 123 entries for the Dairy industry, 80 for the beef industry, 25 in swine, 17 in sheep, and 1 article on the horse industry, which was published in 1932. 8 At the present time, there is an inadequate supply of information available for the operator who is considering expanding or starting a profitable horse business. There are no guidelines available for identifying and quantifying operating costs. If the horse industry is going to move forward in the business place, these operators need the methodology by which to evaluate their facilities, and benchmarks to allow them to determine what types of management decisions need to be made. This study will report which expenses constitute the largest portion of the total expense for operators of boarding, riding, and training stables. It will also report benchmarks, the mean dollars/horse/year that operators are spending in each of the expense categories surveyed. CHAPTER II LITERATURE REVIEW The key to an extended life in the horse industry is increasing profitability, just as in any other aspect of the livestock industry. Although there is limited data available on horse operations specifically, extrapolations can be made from other industries to help calculate the profitability of an operation. The dairy industry is by far the leader in economic evaluation of an operation, and many of these concepts can be applied to the horse industry with a few changes. Profitability of dairy farms can be measured by some key farm business parameters (Galton, 1990). The number of cows per worker, pounds of milk sold per cow per year, pounds of milk sold per worker per year,cost of producing a hundredweight of milk, debt per cow and debt service per cow per year, capital investment per cow, and rate of return on equity per cow. Telfarm is an example of a package created to help operators in the dairy industry to make these profitability calculations, and make informed.management decisions. Telfarm provides a format for a producer to itemize and calculate expenses and incomes in every enterprise in the business. In addition, sectors of the dairy industry publish summaries of this information, to»give operators1a set of benchmarks to use for comparison. This study used these materials to assist in the decisions relative to which expenses to analyze and which 10 to exclude. In addition, opinions were solicited from participants in the industry to formulate a collection of expenses which were perceived to be of value. The sheep industry has also published a handbook, Guidelines for Production and Financial performance Analysis for the Sheep Producers(1992) , to measure the profitability of sheep operations. This handbook includes calculations to quantify feed and pasture expenses and reproductive criteria such as number of lambs per ewe exposed. Studies in the swine industry have shown that there is a decrease in cost of production as operation size increased (Canadian Journal of Animal Science, 1992). This research will try to identify a similar economy of scale in the horse industry. These parameters can all be adapted to the horse industry with a few’ minor changes. In a boarding or training operation, change the income unit from a cow to a housing unit, either a stall or the amount of pasture necessary to house one horse. In a breeding or riding operation, the individual horse becomes the income unit, and sale of offspring or collection.of breeding or riding fees becomes the equivalent of the milk income. Data presently available on the horse industry consists primarily of censuses, and some figures on overall economic impact of the industry in a specific area. The American Horse Council (AHC) reported 750,000 owners and spectators spent 11 over $15 billion in 1984 maintaining horses, or attending horse events. This $15 billion is divided amongst the farm, manufacturing, and service sectors of the economy. The horse industry in 1984 accounted for 16% of the Gross National Product from the Agricultural Sector. In Michigan, $398 million was contributed to the state's Gross National Product. (AHC, 1985) Boarding was the number one income producing facet of the horse industry in Southern Michigan in 1971, followed by breeding and riding respectively. This was based on a survey of recreational horse facilities done in 1971 (Holleran, 1973). Michigan Equine Mbnitoring System(MEMS) conducted a survey in 1991 that indicated that training operations were now leading the income category and boarding had fallen to third. Breeding still ranked second. The horse industry can be approached from two angles, either as a hobby or as a business, and many operations are a combination of the two. A survey conducted by the MSU Cooperative Extension Service in 1971 revealed that only one- third of all horse enterprises hired labor from outside the family, and many of those were only seasonal employees. For this reason, it is difficult to ascertain exact labor expenses. A 1984 study by the USDA and MDA (Michigan Equine Survey, 1984), ascertained that there were 2600 full time employees earning'an average of $8100 annually, and 5200 part- time employees averaging $1800 of income annually. This 12 current study reflects the same difficulty with labor as those conducted previously; Most operators of livestock industries do not report their own labor or the labor of family members, skewing the labor data availabLe. The MEMS 2nd round has tried to deal with this problem by receiving reports of unpaid labor from operators, however, that information is not yet available. There have been many changes in the tax structure dealing with the horse industry in the last ten years, and it is not one of the objectives of this study to analyze these, only to show that improving and quantifying the profitability of the horse operation is of increasing importance. The 1993 Presumption of Profit Publication from the Internal Revenue Service states that breeding, racing, training, and showing operations must show a profit in 2 of seven years. All other operations, including boarding and riding, must show a profit in three of five years to be considered a business, not a hobby. There is a limitless future for advancement of the horse industry, as there is a constant inflow of new participants. Nationwide 4-H programs show that there are more young people involved with horse projects(230,000) than with cattle projects(150,000) (Dyke, 1978) . In 1993, there were 10,682 4-H horse projects in the state of Michigan (Michigan Conference on the Horse Industry,1994). A survey in 1984, conducted by the American Horse 13 Council, reported that 750,000 Americans spent. approximately $15 billion on the horse industry. This money was spent maintaining the estimated 5.2 million horses in this country, and.by spectators attending or betting on horse events. In addition to dollars spent by spectators and owners, the horse industry makes contributions to the economy through jobs, ranging from grooms and trainers, to construction workers to build horse facilities, and.maintenance crews for showgrounds and racetracks. There are also people employed working for tack and feed stores, veterinarians, and publications which interact with the horse industry. Participants in the horse industry also contribute to the general economies of the areas in which they attend shows and other events. A 1991 study of the economic impact of the Virginia Horse Center, estimated that participants at shows and events hosted by the Horse Center spent $9.1 million in Virginia, with $7.7 million of that being in the Lexington- Rockbridge area immediately surrounding the Virginia Horse Center. The majority of this money was spent on food, lodging, and. travel, making it apparent that the horse industry impacts not only the agricultural economy, but the broader economy as well. These figures do not include monies brought into the area through non-show events taking place at the operation. The horse industry in Michigan contributes over 256 million dollars to the Michigan economy each year. These 14 dollars are generated by the 28,700 owners and operators that own or care for the state’s 130,000 horses (MEMS 1991). These owners and operators also realize $122,000,000 of income yearly. The largest percent of this income (34%), is generated by the 1600 riding,training and boarding facilities, with a total of 42 million dollars earned (MEMS 1991). A 1970 study on the horse industry in Michigan reported breeding as the primary income producing facet of the horse industry (Holleran, 1973). The survey conducted in 1972 by the Michigan State University Agricultural Experiment Station, estimated that total investment in horses and facilities exceeded $275 million, and covered over 35,000 acres in Michigan (MAES Report 323) . Only one third of the operators of these facilities employed outside labor, the :majority of them utilized family members as the primary labor source. This survey also projected that by 1985, the horse population in Michigan would grow from 171,000 in 1972, to exceed 400,000. The primary reasons for this growth would be an increase in available leisure time, an increase in family income,and a move toward suburban living (MAES Report 185). The Experiment Station survey cited 3 factors which would discourage the growth of the horse industry, and as of 1994, all of these things have become factors in the horse industry, and agriculture in general. These include concerns relative to pollution of the environment by animal waste, increasing 15 urbanization leading to changing zoning ordinances and a lack of land available to house horses and other livestock, and changes in tax laws. The horse industry in Michigan and around the country was dealt a major blow in 1986 when the Internal Revenue Service made sweeping changes in the taxation status of horse facilities. (American Horse Council data, 1985-88) These changes included the Income Tax Reform and Economic Recovery Act, which changed the taxation status on breeding stock, making them a less appealing investment. Before these tax changes, investors could use breeding operations as tax shelters. The ability to do this was greatly reduced after 1986. This may be partially responsible for breeding falling from its 1970 position as the primary income producing operation. Records of total registrations in the American Quarter Horse and the Arabian appear to reflect this (Table 2.1). TABLE 2.1 NUMBER OF ORIGINAL REGISTRATIONS FROM 1971-1993 Year Quarter Horses1 Arabians2 1971 72,681 2,133 1980 137,009 6,569 1987 147,000 9,465 I 1988 128,000 8,750 I 1990 110,597 6,391 1993 104,876 1,326 n American Quarter Horse Association 2 Arabian Horse Registry 16 A 1984 survey of the Michigan horse industry conducted by the Michigan Agricultural Reporting Service estimated that the horse industry injected over $200 million into the economy, and that revenues in the horse industry were approximately $180 million. The primary use of horses continued to be for pleasure, as it was in the earlier surveys. Total horses in Michigan in 1984 was estimated at 160,000, well below the projection of 400,000 estimated by the Agricultural Experiment Station in 1972. A tremendous drop in number of ponies in Michigan contributes to this general decrease. 1972 projections estimated that the pony population would exceed 61,000, the 1984 survey reported.only 21,000 ponies, a decline of 66%. Concerns relative to the horse industry are not confined to Michigan. The Wisconsin Agricultural Statistics Service conducted a survey in 1992 to quantify the impact of the horse industry on the Wisconsin economy; It is estimated that value of horses, land and buildings was approximately $655 million, and $125 million was spent annually to maintain the estimated 84,300 horses in Wisconsin. Once again, as in the Michigan studies, the primary use of horses was for pleasure. Presently, most available literature on the horse industry deals more with physical management of the horse than with financial and economic management of the facilities. The literature suggests nutritional and exercise regimens, as well as housing and health care recommendations. (Johnstone, 1908) 17 If the horse industry is going to remain competitive and profitable into the 21st Century, operators will have to become more aware of the operating expenses on their operations. In addition, information which will assist operators in making calculations of expected costs will be of value when requesting business loans for starting an operation, or expanding one. This study will provide some benchmarks on selected expenses which may help the operators make some of these decisions. CHAPTER III MATERIALS AND METHODS The data analyzed in this study were obtained through the Michigan Equine Monitoring Service (MEMS), and were collected by the Michigan Agricultural Statistics Service(MASS) , in 1991. The sample was comprised.of two sampling frames, a list frame and an area frame. The list frame was created by collecting names from sources such as: breed associations, telephonetdirectory listings of stables, large equine operations provided by county extension offices,large equine operations from the 1984 National Agricultural Statistics Service (NASS) list sampling frame, farms on the active NASS list frame with equine, and equine owners who attended the 1990 Stallion Show; The Michigan Department of Agriculture also provided lists of equine owners whose veterinarians jperformed Equine Infectious Anemia (Coggins) tests, owners with stallions for service, licensed riding stables, and racetracks. Duplication was removed, and over 13,000 names remained to comprise the list sampling frame. The list sampling frame was stratified by using size of operation,type of operation, and major breed or type of equine. Twenty strata were created and sorted by zip code, and systematic 18 19 random samples were drawn. Sampling fractions varied from 10-100 percent, and the total list frame sample was 2998 owners. The area frame consisted of 338 segments, which are sections of land of approximately one square mile. Data from these segments which did not show up on the list frame were added to the total. Of the 338 segments, 272 had operators with horses on the facility as of June 1, 1990.(Michigan Equine Survey,1991) The area frames were added to be sure to include smaller farms, backyard horseowners, and those who did not have a breed affiliation. The questionnaire (Appendix I) asked operators and owners to rank the level of their facilities involvement in the following activities: Boarding Farm, Breeding Farm, Other Farm, Racetrack, Riding Stable, Private Residence, and Training Stables. More detail on the actual survey technique can be found in the Michigan Equine Survey 1991. (Michigan Equine Survey, 1991) As mentioned earlier, there are some drawbacks in the use of survey data, such as high variability in responses and differences in respondents interpretations of the questionnaire. The advantage of the technique utilized in this study was in the large number of observations obtained. Upon completion of the survey, and deletion of statistical 20 outliers, a database of 2040 operations remained. This study will concentrate analysis on service- oriented facilities, specifically, boarding, riding, and training operations. Most facilities offer a combination of options, such as boarding and training, or lessons, (Holleran, 1973) but this study will concentrate on what was reported as the primary activity. Boarding stables are generally those that provide basic care, such as feed, housing, and turn out if necessary, and can range from full service facilities with all horses stalled, to an operation where the horse lives in the pasture year around. Training stables offer the same basic care, although horses are generally stalled, and in addition, the horse will be ridden or otherwise taught by the operator. Training stables cover the gamut from places where the horse lives for 30-60 days and is started.under saddle, to facilities where the horses live all year around, and are shown at varying levels of competition. Riding stables are defined as those where horses can be rented at a predetermined rate for trail riding, or other equestrian activities. The original database was also reduced to exclude any facilities where the major activity is racing either Standardbreds or Thoroughbreds. These enterprises were excluded because there is possibly a larger variation in expenses such as feed and health because of the physical stress these animals undergo. It is believed that the increased energy expended, and the more strenuous physical 21 exercise would create sufficient differences for racing to merit being analyzed separately. Facilities were also deleted which reported a total expense of zero or less, and those which reported a total expense of greater than $40,000 per horse per year. Facilities were also divided into four size categories. Size category 1 had 1-2 horses (very small), size category 2 housed 3-9 horses (small), size category'3 housed 10-19 horses (medium), and size category 4 housed 20 or more horses (large). Although size category 1 represents 18% of the horse population in Michigan (MEMS 1990), these facilities were eliminated from this analysis, based on the assumption that these were not operated as income producing enterprises. These operation sizes were arrived at based on the accumulated ideas of people with years of experience in the horse industry. As stated earlier, the objectives of this study are * identify differences in expenses based upon operation type,regardless of operation size *identify differences in expenses based upon operation size, regardless of operation type *identify differences in expenses by the interaction of operation size and operation type 22 Using Statistical Analysis Software (SAS), the condensed data set was used to calculate means and ranges on the following expenses on a $/horse/year basis: TABLE 3.1 EXPENSES SELECTED FOR ANALYSIS Board paid Training fees paid Breeding fees paid Health costs Feed(purchased and raised) Insurance (equine related only) Farrier Travel and Lodging (equine related only Advertisement Utilities Capital Improvements Maintenance and Repair lIBedding Tack Rent/Lease (land and/or Taxes (Property tax buildings) excluding residence) Professional fees Miscellaneous Labor Expenditures Other Equipment Purchases Total Expenses Feed expenses were defined as the value of any feed purchased or raised for use on the operation. Labor expenditure was how much was paid for labor. This does not include value of free or bartered labor contributed by family members, or any other people. Total expenses consisted of the summation of all expenses on the questionnaire. After the first summarization of these expenses, obvious outliers were removed from the data set. After removal of the outliers, the data were again processed to obtain the most 23 accurate estimates of each of these expenses as possible. The raw data had heterogeneous variance, so all data were transformed to natural log to increase homogeneity. All analysis was conducted on the logged data. The data were also appropriately weighted based on sampling methods discussed earlier. The data were returned to normal values for reporting of means. Adjusted-log means are also reported. All analysis was done on the adjusted-log means. The first test run was an Analysis of Variance (ANOVA). ANOVA compares actual mean squares to the value that would be expected in the situation of a null hypothesis, which states that there is no difference between the expected and the actual value. If the null hypothesis is false, and there is a difference present, the numerator increases, increasing the F value. A large F value rejects the null hypothesis. To determine significance of any difference present, the P value is used. The P value estimates the probability of getting an even larger F value if the null hypothesis is true. The lower the P value, the greater confidence that rejection of the null hypothesis is correct. Because this study contains unbalanced data, the ANOVA must be run with a linear model. The linear model uses a matrix which weights the number of observations meeting the specified restrictions. This is necessary with unbalanced data to avoid assigning the same importance to means containing different numbers of observations. A mean with 50 24 observations should be weighted differently than a mean with one observation. The following linear model was used by this study: with T representing type of operation, S representing operation size, TS representing the interaction of operation size and operation type, and E representing error. The significance of the model was tested for each of the expenses in Table 3.11 Any expense where the model was not significant at P < .10 was eliminated from further consideration because the larger the P value, the greater the probability that the null hypothesis is correct. If the model is not significant any comparisons made with that model are invalid. The ANOVA identified which of the expenses on Table 3.1 showed differences by operation type, size category, or the interaction of operation type and size category. If P> .10, the relationship was considered insignificant, and the expense was excluded from further statistical analysis. If the expense was insignificant by the ANOVA, any level of significance identified by later test would be invalid. This test was followed by calculation of adjusted-log means that would have been expected had the data been balanced. P values were then calculated from these adjusted- log means to identify significant interactions between combinations of operation type and operation size. The study 25 concentrated analysis on the five expenses that contributed the highest portion of the total expense on boarding, riding and training facilities. This was determined by ranking the adjusted-log mean for each expense which was significant by the interaction of operation type and size category. Miscellaneous and other expenses were eliminated from further analysis as they were judged to be too unspecific to provide valid comparisons. Any expense which was in the top five of 2 operation types is reported, as is labor. Labor was included because it was believed that this expense was important. The questionnaire asked for expenditures on labor, an area where it would be anticipated that operation type and size may have a large affect. Labor is a large expense on most businesses and farms, and would be expected to have the same influence on horse operations. CHAPTER IV RESULTS i!!i-§ 0 ’_SU O M 1 L. 1'. _ ;V_- The 1991 Michigan Equine Survey, which provided the data base for this study, estimated the Michigan equine population to be approximately 130,000 head. This is a decrease of 30,000 from the 1984 inventory (Michigan Equine Survey, 1984) . The 1984 Michigan Equine Survey did not use an area frame, it used only a list frame. The list frame consisted of a list of names of participants derived from veterinarians, breed associations and organizations, etc. The area frame was derived by randomly surveying 1 mile square areas in the state. This difference in methodology may have contributed to some of the population differences between the 1984 and 1991 surveys. American Quarter Horses, Standardbreds, Arabians, and Thoroughbreds were the most populous breeds in Michigan. American Quarter Horses showed the sharpest decrease in population from 1984 to 1991, from 41,500 to 29,000. Standardbreds dropped from 19,000 in 1984 to 17,000 in 1991. Arabians and Thoroughbreds increased by 1,000 and 900 respectively. The other major decrease occurred in ponies, from a total of 21,000 in 1984 to 13,000 in 1991. 26 27 The 1991 Michigan Equine Survey used slightly different size categories than did this study. TABLE 4.1 ESTIMATED DISTRIBUTION OF OPERATIONS IN MICHIGAN BY OPERATION SIZE1 NUMBER OF OPERATIONS ,‘ f OPERATION SIZE : 1-2 horses 14,700 g 3-9 horses 11,500 ; 10-29 horses 2150 n ’30 + horses .7 350 I 1 MEMS, 1991 The majority of the horse population was located in urban areas, near human population. The top four counties in horse population were Oakland, Washtenaw, Livingston, and Jackson. Population estimates per county are not presently available. Registered horses constituted approximately 60% of the total population. This number is the same as in 1984. The breeds with the highest percent registered were Thoroughbreds and.Arabians at 90%, Morgans at 80%, and Standardbreds at 79%. The Michigan Equine Survey also determined the primary use of horses to be for pleasure,42.3%. Pleasure was also the primary use of horses in the 1984 survey, 52.2%. In the 1991 survey, breeding replaced showing as the second most common use of horses. Racing, both flat and harness, was the primary occupation of 12.3% of Michigan horses in 1991. 28 TABLE 4.2 INCOME AND EXPENSES BY OPERATION TYPE1 : 7 7 --~— ~- -—— - —- m i OPERATION’ NUMBER OF INVENTORY 1990 1990 TYPE OPERATORS JUNE INCOME EXPENSES ,,l ‘ 1,1991 DOLLARS DOLLARS Private 16,500 45,000 11,000,00 61,000,00 Residences 0 0 Other 9,000 30,000 7,000,000 23,000,00 Farms 0 Breeding 1,500 23,000 22,000,00 49,000,00 Farms 0 0 Training 900 11,000 26,000,00 39,000,00 Stables 0 0 Boarding 500 12,000 12,000,00 17,000,00 Stables 0 0 Riding 200 3,000 4,000,000 14,000,000 Stables Other 100 6,000 6,000,000 5,000,000 Operations Owners NA NA 34,000,00 58,000,00 with all 0 0 Horses Boarded Elsewhere 28,700 130,000 122,000,0 256,000,0 00 00 — ~—4Mflfl 1 MEMS 1991 29 TABLE 4.3 TOTAL DOLLARS SPENT BY OPERATORS OF MICHIGAN HORSE OPERATIONS1 EXPENSE TOTAL DOLLARS AVERAGE DOLLARS jFeed 40.0 308 : Equipment Purchases 36.0 277 i . Equine Purchases 35.0 269 Training Fees 17.0 131 Health 15.0 115 . Capital Improvements 15.0 115 | Boarding of Equine 14.0 108 A Tack 11.0 85 1 Farrier 9.0 69 l Travel & Lodging 8.5 65 % Taxes 8.0 62 Labor 8.0 62 J Miscellaneous 7.0 54 Maintenance/Repair 6.5 50 " Breeding Fees 6.0 46 ‘ " Bedding 5.5 A 42 I "Insurance 4.0 31 E Utilities 4.0 31 Professional Fees 2.5 19 Advertisement 1.5 11 Rent/Lease 1.5 11 Other 1.0 8 TOTAL 256.0 1,969 i ‘ MEMS, 1991 30 TABLE 4.4 DISTRIBUTION OF OPERATIONS IN THIS STUDY l Boarding Riding Training TOTAL 3-9 horses 29 8 24 61 10-19 horses 35 11 24 70 20 + horses 36 24 34 94 TOTAL 100 43 82 225 A comparison with the estimates of operation numbers in Table 4.2 shows that twenty percent of the boarding stables, twenty-one percent of the riding stables, and nine percent of the training stables in Michigan participated in this analysis. §IAIL§TLQAL_QYEBYIEH Results of the ANOVA on the expenses in Table 3.1 are reported in Table 4.5. A P < .10 was required for significance. Significant values are in bold type. Table 4.5 shows that the model for capital improvement was not significant, eliminating it from further analysis. Total expenses, insurance, advertising,bedding, tack, farrier, travel, miscellaneous, and.health expenses were significantly different by Operation type. Total expense, labor, tack, rent, other, and equipment purchases were significantly different by size category. Feed, total expense, insurance, advertising, utilities, maintenance, bedding, tack, taxes, farrier, miscellaneous, other, professional fees, health, and equipment purchases were significantly different by the interaction between Operation type and size category. 31 mufiHmsw nu each a“ oH.ve page. cane. cane. was”. mucosm>oumeA Amadeus Hoes. smsa. mesa. sees. meanness Hues. ease. ”Hem. nose. nauseousm newsmaswo some. menu. mmsm. undo. oueon deco. mews. nose. Hooo. season «one. mean. mass. ”coo. some HmcoemmoOoum eene. enee. Heme. some. umsuo some. Amen. Hoes. «coo. usomsuaamoues seem. Ones. some. egos. uses nmom. mean. ease. deco. Hm>suu deco. mmmfl. Hoes. Aces. Susanne ease. meow. «man. sees. nauseous nose. seem. smem. once. noun» «coo. memo. mace. nooo. some 4666. seem. ease. Hose. museums Hose. mesa. seed. Hoes. oosecmusflms smme. same. mean. eneo. nosed ”see. same. Hmme. "coo. manusasus «Hoe. mafia. «coo. ”coo. menueuuo>ee ammo. HHHm. «nee. coco. oucuusuea «coo. eeae. «use. Aces. musmmxo sauce made. mess. move. Hose. some use 03m .. EEO Commune 03m 0&5 .8330an deco: 033.25 09 m0 mHmwflflfi mmahd mmmzmmxm QmBUmAmm m0 mMDQ¢> m m.¢ mags 32 TABLE 4.6A RANKING OF EXPENSES ON BOARDING STABLES1 BY ADJUSTED-LOG MEANS Adjusted-log Mean Actual mean (log$/horse/year) ($/horse/year) Total expense 6.92 1793 Feed 5.24 254 Health 3.62 79 Bedding 3.34 66 Maintenance 3.23 85 Taxes 3.14 84 Farrier 3.14 42 Utilities 3.01 51 Tack 2.94 50 Insurance 2.67 54 v I Equipment 2 . 51 368 Purchases l Professional 2 . 07 70 Fees Labor 1.83 97 lAdvertising 1.58 11 I 1 n=100 33 TABLE 4.6B RANKING OF EXPENSES ON RIDING STABLES1 BY ADJUSTED-LOG MEANS l Adjusted-log Mean Actual mean (logS/horse/year) ($/horse/year) Total Expense 6.84 1487 Feed 5.37 276 a Health 2.54 47 Utilities 2.33 47 Tack 2.32 53 Taxes 2.29 73 Maintenance 2.25 47 Farrier 2.24 43 Insurance 2.08 76 I Equipment 2.04 171 Purchases Advertising 1.8 24 Labor 1.73 45 Bedding 1.46 39 Professional 1.05 14 Fees =43 34 TABLE 4.6C RANKING OF EXPENSES ON TRAINING STABLES1 BY ADJUSTED-LOG MEANS Adjusted-log Mean Actual means (logS/horse/year) ($/horse/year) Total Expense 7.48 3040 Feed 5.39 283 Health 4.12 108 Farrier 3.77 84 Tack 3.64 115 Bedding 3.41 78 Maintenance 3.41 113 n Equipment 3.17 930 | Purchases Utilities 2.94 68 Taxes 2.78 64 I Advertising 2.43 26 Labor 2.00 91 Insurance 1.93 47 Professional 1.6 28 Fees 1n=82 35 Tables 4.6A-C show that feed, and.health were the top two expenses for all operation types. Bedding and tack were also found in the top five on two or more operation types, qualifying them for further examination. Although taxes met this requirement, this expense was not analyzed further as it is beyond the control of the operator. Although it was not a top ranking expense in this survey, labor was added as it is a generally a major expense in most businesses. The means reported on Tables 4.7 through 4.11 are the All statistical actual means and the adjusted-log means. analyses were performed on the adjusted-log means. Superscripts identify which comparisons are significantly different at P < .10. DIFF E O RAT 0 TABLE 4.7 ACTUAL1 AND ADJUSTED-LOG MEANS2 OF SELECTED EXPENSES COMPARED BY OPERATION TYPE3 Boarding Riding Training n=100 n=43 n=82 === ~- Health 79 47 108 Actual 3.62a 2.54° 4.12b Adjusted-log Tack 50 53 115 Actual 2.94a 2.32a 3.64b Adjusted-log Bedding 66 39 78 Actual 3.34a 1.46b 3.411 Adjusted-log Total 1793 1487 3040 Actual Expense 6.92a 6.8481 7.48b 1 $/horse/year 2 log$/horse/year 3 Values in rows with different superscripts denote differences, P < .10 36 Table 4.7 shows that there are differences among the selected expenses based on Operation type. Feed and labor are eliminated from this comparison as they were not significant by Operation type when the ANOVA was performed. Comparisons of costs by Operation type are based on the probability of a difference occurring between the adjusted-log means of the selected expenses. In a comparison between boarding and training operations, training Operations have higher expenses on a $/horse/year basis in both total expenses and health costs (Table 4.7) There is no difference in bedding‘ between boarding and training stables, when all 3 size categories are considered. When comparing boarding and riding operations, boarding has higher costs/horse/year in health and bedding expenses (Table 4.7). There is no difference in total expenses. Training Operations demonstrate a higher annual cost in total expenses, health, bedding, and tack expenses than do riding Operations (Table 4.7). When training Operations are compared with either boarding or riding operations, they show a higher annual cost. Riding operations have the lowest annual costs in expenses where a difference exists. 37 TABLE 4.8A RANKING OF EXPENSES ON SMALL OPERATIONS1 BY ADJUSTED-LOG MEANS I Total Expenses “Adjusted-log Mean Actual Mean 1n=61 7.47 3573 Feed 5.38 296 Equipment 3.72 1295 Purchases Health 3.64 124 Farrier 3.29 86 ;Tack 3.22 111 1 Maintenance 3.03 129 ; Utilities 2.95 86 [Bedding 2.93 95 ITaxes 2.83 90 iInsurance 1.83 46 . Professional Fees 1.82 88 Advertising 1.54 23 ‘ Labor 1.02 93 I 38 TABLE 4.8B RANKING OF EXPENSES ON MEDIUM OPERATIONS1 BY ADJUSTED-LOG MEANS _ Adjusted-log Means Actual Means 1 n=70 Total Expenses 6.96 1448 Feed 5.27 202 Tack 3.33 59 ' Health 3.30 54 Farrier 3.00 39 Maintenance 2.86 47 Bedding 2.77 47 Utilities 2.68 31 | Taxes 2 . 66 84 Insurance 2.52 62 Advertising 2.15 13 Labor 2.15 128 Equipment 2.02 137 Purchases Professional Fees 39 TABLE 4.8C RANKING OF EXPENSES ON LARGE OPERATIONS1 BY ADJUSTED-LOG MEANS ‘- Adjusted-log Means Actual Means 1 Total Expenses 6.83 1158 Feed 5.34 257 l Health 3.35 55 Maintenance 3.02 54 l Farrier 2.85 33 Taxes 5.72 37 I Utilities 2.66 34 Bedding 2.53 36 Labor 2.40 66 Tack 2.36 29 Insurance 2.34 36 Advertising 2.13 15 Equipment 1.99 171 1 Purchases Professional Fees 40 Tables 4.8A-C show that feed is the top individual expense in. all size categories, just as it. was in all Operation types. Health, the second expense by all Operation types, was in the top three expenses of all size categories. The means reported Tables 4.9-4.12 are the actual means and adjusted-log means. All statistical analyses were performed on the adjusted-log means. Superscripts identify which comparisons are significantly different at P< .10. IFFERENCES Z GORY TABLE 4.9 ACTUAL1 AND ADJUSTED2 MEANS OF SELECTED EXPENSES COMPARED BY SIZE CATEGORY3 Small Medium Large 3-9 horses 10-19 horses 20+ horses Means n=6l n-7O n=94 Labor 93 128 66 Actual 1 . 01° 2 . 14ab 2 . 40b Adjusted-log Tack 111 59 29 Actual 3 . 21' 3 . 32° 2 . 35b Adjusted-log Total 3573 1448 1158 Actual Expense 7 . 47' 6.96b 6. 32b Adjusted-log E 1 $/horse/year log$/horse/year Value in rows with different superscripts denote differences, P < .10 Table 4.9 shows the differences in each selected expense by size category. Feed, health, and bedding were eliminated from this comparison because they were not significantly different by size category when the ANOVA was performed. Superscripts are used to show which comparisons showed a difference at P < .10. These differences were detected by 41 calculating the probabilities of differences using the adjusted-log means. The comparison between small and.medium categories shows that total expenses were higher for small operations (Table 4.9) . Labor and tack showed no differences between these categories. The comparison between small and large Operations reveals differences in total expenses, labor, and tack expenses. Small Operations had higher total expenses, and tack expenses, while large Operations had a higher labor expense. The comparison of medium and large Operations revealed a difference in tack expenses, with the expense for medium Operations being higher (Table 4.9) . Total expenses and labor showed no difference between medium and large Operations. This analysis shows that although there is no difference in total expenses between medium and large Operations, both medium and large Operations have a lower total expense on a $/horse/year basis than do small operations. T O S ETWE N N PE AND I CA Tables 4.10-4.12 show the differences in selected expenses between different sized Operations of the same type. Tables 4.13-4.15 show the differences in selected expenses between different operation types of the same size. The same expenses were considered that were used in the analyses by Operation type and size category. Actual means 42 and adjusted-log means are reported, with superscripts to identify differences at P < .10. Calculations Of the probability of differences existing were performed on the adjusted-log means. TABLE 4.10 ACTUAL1 AND ADJUSTED-LOG MEANS2 OF SELECTED EXPENSES COMPARED BY SIZE CATEGORY WITHIN BOARDING STABLES3 Small Medium Large Means 3-9 horses 10-19 horses 20+ horses _ _ n=29 n=35 n=36 . Feed 334 174 267 Actual . 5 . 69° 4 . 69b 5 . 34a Adjusted-l . . ' Health 125 42 50 Actual 4.51‘1 3.061’ 3.291’ Adjusted-log . Tack 60 43 41 Actual . 3 . 48' 2 . 651” 2 . 68" Adjusted-lo 1 1 Bedding 97 38 59 Actual ‘ 4 . 10° 2 . 61'” 3 . 32"c Adjusted-lo . Total 2518 1175 Actual Expense 7.50'1 6.24b Adjusted-log 7 —~— -—---—— um ——- ———— -———-— —-— —————-— - -— 1 S/horse/year 2 log$/horse/year 3‘Values in rows with different superscripts denote differences, P< .10 Table 4.10 shows that there are differences in selected expenses on a $/horse/year basis among boarding stables based on size. Labor was eliminated from this comparison as it was not significant by the interaction of size category and Operation type when the ANOVA was performed. Small boarding stables had higher annual costs than did medium boarding stables in all selected expenses . Small stables were higher in total expenses, feed, health, bedding, and tack. Health expenses were higher on small boarding Operations 43 than on large operations (Table 4.10). There were no differences between large and small boarding operations in total expenses, feed, bedding, and tack expenses. Large boarding stables had higher feed and total expenses than did medium boarding stables. Health, bedding, and tack expenses showed. no