Parimutuel wagering is conducted in one form or another in 32 States that contain approximately 80 percent of the Nation's population. Most parimutuel wagering takes place at horse tracks-thoroughbred, harness, and quarter horse. Dogracing is the second most popular parimutuel event, while jai alai, the third, generates approximately one-eighth of Florida's gambling income and a barely measurable percentage of Nevada's. The newest form of parimutuel wagering is off-track betting, which is an extension of on-track betting and operates according to the same general principles as on-track

betting. Although it is conducted only in New York State and Connecticut, OTB is growing in importance as a new and potentially major form of legalized gambling.

This section examines the various aspects of the parimutuel industry-its history and development, its operation, the revenue it produces, and the problems that confront its principal components. The section also contains a brief description of the State regulatory bodies that govern parimutuel wagering and of the industry's many independent trade associations.

HORSERACING. Horseracing-commonly called the "sport of kings"-has evolved over the centuries from its aristocratic origins into a widely popular pastime. Crowds for racing meets are still increasing; racing is the most well attended sport in America.

Although the first recorded mounted horserace occurred in Greece in the seventh century, B.C., horseracing in one form or another appears to be at least 6,000 years old. Flat racing-in which a rider is mounted directly on the horse, not in a rig drawn by the animal-is about 3,000 years old. Racing gained its regal reputation and its greatest impetus in England, where it became a preoccupation of the monarchy from about the 12th to the 17th century.

The scientific breeding of thoroughbred horses is usually credited to the Duke of Cumberland in England in the 18th century. The first known thoroughbred horserace occurred in England in 1764.

The three principal forms of horseracing in the United States today are harness, thoroughbred, and quarter horse. Less common are Appaloosa, Arabian, and Paint horseracing. Harness racing is believed to have preceded flat racing by some 3,000 years, dating back to the era when horses were used as beasts of burden and before riders mounted the animals directly. The chariot races of ancient Rome were a fairly well-developed phase in the history of the sport. Today, although trotting has been called "the most typically American of all horse sports," 23 its role in the racing economy is secondary to that of thoroughbred racing. In 1975, there were fewer thoroughbred than harness racing days but the thoroughbred handle was $2 billion higher than the harness racing handle (see table 5-20). Nineteen States currently have authorized parimutuel wagering at harness races. Nighttime harness racing, a factor in the continuing success of the sport, began at Roosevelt Raceway in 1941. There are a number of reasons why American racing is dominated by the thoroughbred. Its traditional appeal to those in a position to make important political decisions favoring the sport may be a factor in its proliferation. In the early Colonial days, concern over the

lack of quality horses in America prompted colonial governors to sponsor races as a means of identifying the fastest horses for selective breeding. The early racesattended primarily by the American "gentry"-eventually evolved into organized sporting events whose appeal transcended class lines.

One curious aspect of thoroughbred racing is that no one knows exactly what constitutes a thoroughbred. Today, in order to qualify as a thoroughbred, a horse must be registered with the New York Jockey Club. Only those horses may be registered whose sire and dam are registered, and, ostensibly, the lineage of all thoroughbreds must be traceable genetically back to the three prototypes of the breed, oriental stallions known as Godolphin, the Byerly, and the Darley. Anatomically, the thoroughbred is distinguished from other breeds of horses by its greater height and longer legs.

The quarter horse might be called the country cousin of the patrician thoroughbred. The quarter horse-whose name refers to the quartermile sprint it runs on the trackwas developed in colonial Virginia by breeding native English horses to horses of Spanish ancestry, producing a swift hybrid for the short distances required for it. Unlike the thoroughbred, the quarter horse is bred not merely for racing and breeding, but for more utilitarian purposes as well, such as farm work and even transportation. In 1974, more quarter horses-some 1 million-were registered than all other breeds combined.

Despite the prevalence and versatility of the quarter horse, it plays a relatively minor role in the parimutuel industry. Quarter horse meets are often rural activities run at State and county fairs and other community functions. Parimutuel wagering on quarter horse races takes place in 16 States, usually in mixed races on thoroughbred programs. There is no parimutuel wagering conducted at about a third of all quarter horse races. In 1975, the total parimutuel handle for quarter horses was only $140,345,856 (exclusive of the handle for mixed programs of thoroughbreas and quarter horses). In the coming years, however, the growing shortage of thoroughbreds may

foster the growth of quarter horse racing, as more track owners begin using quarter horses to fill gaps in their thoroughbred racing programs.

The formal debut of horseracing in the American colonies took place in 1665, when a New York governor sponsored racing meets at the Newmarket track on Long Island. Although wagering became a factor in the sport almost immediately, horserace betting in the United States remained largely an upper class activity for about 200 years; its chief participants were the breeders and owners who supplied the animals. But by the mid-19th century the general public had developed an appetite for the sport and for the betting that seems indigenous to it.

To meet the needs of racing's expanding market, and to generate still more fans, racing entrepreneurs began to build huge new track facilities capable of holding tens of thousands and offering a festive, circus-like atmosphere for the players. New York's Belmont Park, which opened in 1905, was the prototype of these tracks; over the years, it was to be matched in scope and clientele by such parks as California's Santa Anita (1934) and New York's Aqueduct (1959), the Nation's largest.

As betting increased among racing fans at the new tracks, a new gambling figure emerged: the racetrack bookmaker. The bookie rapidly became the dominant figure in horse betting. Usually a knowledgeable habitué of the track, the bookmaker was the agent against whom the layman bettor wagered his own expertise and hunches. The bookie, was, in effect, the bettor's competition; with his insider's tips, his greater familiarity with the mechanics of the wager, and his frequently unscrupulous manipulation of jockeys and trainers, the bookie held the clear advantage.

The bookmaker's position in horse betting was weakened by the invention of the parimutuel system, in 1865, by a Frenchman. Parimutuel betting-now the dominant legal system in the field-offered a number of attractive innovations for the average bettor. It allowed him to bet against his peers, the other amateur bettors, instead of the professionals, the bookmakers. By allowing payoffs on second and third place finishes, it increased the bettor's chances of getting some return on his investment.

And by using machines to calculate odds, issue tickets, and provide payoffs, parimutuels went a long way toward eliminating the elements of corruption, fallibility, and arbitrariness that had often marked the bookmaker's trade. First used successfully at the 1908 Kentucky Derby, parimutuel betting is now legal at more than 150 racetracks in 32 States, and is the most popular form of legal gambling in the Nation. Contributing to its growth has been the introduction of long-odds bets that are called "exotic" bets: the exacta, the quinella, the trifecta, etc. Many track fans dislike exotic bets. One track owner said:

I consider normal betting on just one race at a time as wagering. The bettor tries to make an intelligent choice based on his knowledge of the horses, jockeys, and track conditions. The doubles and quinellas, on the other hand, are gambling-a matter of pure luck.

Nonetheless, exotic bets, no matter how "unscientific" they may be, remain popular.

JAI ALAI. Actually, the exacta and quinella are not the invention of the horseracing industry but of jai alai, which, along with dogracing, is one of only two sports to share a parimutuel betting system with horserace tracks. Jai alai (which means "merry festival") has been around since the 17th century, when it was devised by the Basques, but it has found favor as a professional sport among Americans only recently. Parimutuel betting on jai alai was legalized in Florida in 1935; only Nevada and Connecticut so far have followed Florida's example. Rhode Island has legalized the game but has not yet begun its operation. In the more than 40 years since jai alai has been regulated by the Florida Racing Commission, there has never been a charge of a fix, a fact jai alai enthusiasts claim make it the cleanest gambling game in the world.

DOGRACING. Although there are accounts of the use of greyhounds for sporting events in ancient Rome, the greyhound's principal historical role was that of a hunter whose speed and stamina were unmatched by any other canine breed. The earliest greyhound racing was a competitive version of the hunt. Two greyhounds would be released together in pursuit of a hare, and the race would be judged on the basis of the dogs' performance

during the chase. Called "coursing," this type of greyhound racing reached the height of its popularity in England during the reign of Queen Elizabeth I, when it first began to be called "the Sport of Queens."

The modern version of greyhound racing in America evolved from a coursing event held in South Dakota in 1904. The sponsor of the event, Owen Patrick Smith, developed an immediate enthusiasm for the sport along with a strong distaste for the killing of hares. Smith spent the next 15 years testing and perfecting a mechanical lure. These early devices were unwieldly contraptions that often would slam into a wall, fall off the track, or stop suddenly,

causing a pileup of dogs and machine. These mechanical difficulties were resolved with the invention of a folding arm that made the lure disappear while allowing the vehicle to continue around the track. But even with this new technology, greyhound racing did not achieve instant popularity. Many of the early races were conducted by unscrupulous fly-by-night promoters who gave the sport an unsavory reputation. Eventually, however, introduction of the parimutuel system of wagering along with more effective regulation fostered public acceptance of greyhound racing as a legitimate form of entertainment.

The essence of the parimutuel system of betting is that bettors wager against one another instead of against a bookmaker. Of the total amount wagered on a particular race, approximately 82 to 85 percent is returned to winning bettors in the form of payoffs on the three winning horseswin, place, and show-and on combination (exotic) bets where they occur. The minimum wager per race is $2, but larger bets in units of $5, $10, $50, and $100 also may be placed. The amount of money paid to winning bettors is determined on the basis of the total amount wagered on each horse or dog (or player in the case of jai alai). If a particular horse or dog is heavily favored during the course of betting, the payoffs on that animal if it wins will be much smaller than the payoffs on a winning horse or dog on whom little was wagered. For example, if a total of $50,000 is bet on a particular race and $10,000 of that amount is bet on a particular horse, the payoff odds on that horse will be 4-1; that is, for each dollar bet on this horse, $4 has been bet on other horses. The odds on place, show, and exotic bets are not calculated in advance because too many horses and combinations are involved. The total amount bet on place bets is divided among ticketholders whose horse came in first or second; the amount wagered on show bets is divided among ticketholders whose horses came in first, second, or third.

Wagering computations are accomplished by a totalisator (tote) machine, which adds bets over and over again during the course of betting. Every 60 to 90 seconds the tote flashes the new betting totals and odds for each animal on a large toteboard. During the approximately 20 minutes before each race when betting occurs, the odds on each contestant may change dramatically. The machines contain a number of features designed to minimize the potential for parimutuel fraud or machine malfunction. These features include coded ticket paper and duplication of all critical functions by two computers working independently of one another. The most sophisticated of the modern tote machines not only compute and post odds but also maintain complete records of the betting activity on each race as well as postrace information on payoffs and uncashed tickets. Totalisators also are used in jai alai and for off-track betting and can be modified for use in lotteries and sports betting.

The Takeout

The takeout is the portion of the parimutuel pool that is not returned to winning bettors. The amount varies among the States but generally ranges from 15 percent to 1percent of the handle. The takeout is divided among the horse or dog owners' purses, the track, and the State in which the racing is conducted. The apportionment of the takeout among the three entities also varies from State to State. In some States the takeout on combination bets exceeds the straight bet takeout. In New York and New Jersey, for example, the takeout is 25 percent on threehorse combination bets. Often, the State receives half of the total takeout, and the remaining half is divided equally between the track and the horse owners. Purses in dogracing average 12 to 13 percent of the total takeout compared to about 25 percent for horseracing purses.

Three recent research efforts suggest that current takeout rates are too high. The first is an article based on data for New York tracks prior to the introduction of OTB and exotic wagering. The author shows that the takeout rate yielding the highest revenue to the State in 1969 would have been 14.88 percent rather than the 17.16 percent that actually prevailed. He also estimates that this reduced takeout would have increased revenue to the State and tracks by $1.1 million.

The National Gambling Commission also has undertaken research on this question. One study, 25 using data on per capita handle in the individual States for the years 1950 to 1974, estimates the optimal takeout rate at 12 to 13 percent. Another study, 26 using a more complicated model to explain handle at the New York tracks, places the optimal takeout rate at 12 to 14 percent. Although these efforts cannot be considered as absolutely conclusive, they do suggest that there has been some confusion between high tax rates and high tax revenue, and they underscore the need for more detailed research in this area.

In addition to the takeout, there is another, smaller portion of the parimutuel handle that is not returned to winning bettors. That money-called breakage-is the amount of each winning payoff that exceeds multiples of

5 or 10 cents. For example, if the payoff to a winning bettor is $5.67, the bettor will receive only $5.60, with 7 cents retained as breakage. Some States have provisions for a 5-cent breakage that only applies in specific cases. If, for example, the winning horse in a particular race is so heavily favored in the betting that there is not enough in the parimutuel pool to pay the winners (called a minus pool), the breakage may be reduced to 5 cents in order to provide additional money for the winners. Thus, on a payoff of $2.27 a 10-cent breakage will yield the winning bettor $2.20, while the 5-cent breakage will yield $2.25. The distribution of the breakage monies varies among the States. Some States retain the entire sum; others return all or part of it to the tracks, and, in a few cases, a portion of the breakage is used to supplement purses, for breeder and owner awards, or for improvement of backstretch facilities for track personnel-jockeys, trainers, etc.

Finally, for every race there are some people who fail to cash in their winning tickets. The money from uncashed tickets goes either to the State, to the track, or is divided according to a specific formula. In addition to their share of the parimutuel handle, the States may collect money through taxes on track admission, parking, and from track and occupational licenses (discussed below).

Parimutuel Revenue

Almost half of the $19 billion handle from legal gambling in 1975 consisted of parimutuel wagers at horseraces and dograces and jai alai frontons. From the States' point of view, the most important aspect of this wagering is the money it generates for their overburdened treasuries. New York is the major parimutuel State; in 1975, it earned from parimutuel and related incomes $162.5 million. Table 5-21 shows State revenues from parimutuel wagering on horseraces, dograces, and jai alai (excluding OTB) from 1970 through 1975. The figures

listed in table 5-21 include the State's share of the parimutuel takeout plus sums collected through track licenses, occupational licenses, breakage, admission taxes, and miscellaneous fees.

The increased revenue to the States from racing reflects increases in the overall attendance rate, in handle, and racing days. Inflation and the declining value of the dollar have contributed to an increase in the average per capita bet as well. Tables 5-22, 5-23, and 5-24 show the increases in handle and racing days, attendance, and per capita bet, for horseracing and dogracing from 1965 to 1975.

The number of racing days has increased in all States except four. The States have sought to expand the parimutuel revenue base as one means of meeting their spiraling revenue requirements. Paralleling the expansion of horseracing and dogracing has been the growth of other forms of legalized gambling such as lotteries and off-track betting, and movements to legalize forms of gambling such as sports betting and casinos.

Total racing attendance, although increasing, has not kept pace with the rate of growth in racing days. As a result, the average daily attendance rate has decreased by 23 percent for horseracing and 1 percent for dogracing in the past 11 years.

The significance of the increasing racing dates and decreasing daily attendance is difficult to gauge. There is no evidence that the popularity of racing is declining. The increases in total attendance and handle would seem to indicate that the opposite is true. Those within the racing industry are concerned, however, that the State's continuing desire for increased gambling revenues may one day cause the racing industry great harm. As will be discussed later in this section, the growth of racing dates ultimately affects the number and quality of horses and dogs avaliable to race, the performance of individual horses and dogs, and the profitability of racing to horse, dog, and track owners.

The National Gambling Commission survey of American gambling attitudes and practices revealed that 14 percent of the adult population-about 20 million people-bet on horseraces at the track in 1974. Four percent of the adult population—or 5.8 million people-bet on the dograces that year (table 5-25). Proportionately more men than women bet, more Northeasterners than Southerners. Betting at both horseraces and dograces tended to increase as income and educational levels increased. Betting participation peaked among those aged 25 to 44 and then began to decline. Particularly sharp reductions occurred after age 65. Betting at horseraces was three times more prevalent than betting at dograces, the most obvious reason being that there are more horse tracks than dog tracks in the country.

Additional highlights of the survey findings follow:
Family Income and Betting (table 5-26). From the

<*andpoint of social policy, an important question is whether the percentage of income wagered by individuals rises or declines as income increases. As shown in table 5-26, families earning under $5,000 a year spend a higher percentage of their income on horseraces than those earning more money. Although this makes horseracing a regressive form of gambling, it is less regressive than, for example, lotteries and numbers, which are particularly popular among low-income people.

The cost of gambling to any group is not the amount wagered, but the net loss to players-that is, the amount taken out and retained by the operators. Another way of measuring regressivity, then, is to compute the takeout as a percentage of income. The figures show that, as a source of revenue for the government, on-track horserace betting, like most