As has been chronicled on several occasions in Thoroughbred Times, beginning with our series on unsoundness in the American Thoroughbred in 1995, contemporary Thoroughbreds make, on average, about half as many career starts as they did in the 1950s. We have always pointed out that there are several factors that undoubtedly contribute to the reduction in average career starts for the American Thoroughbred from 34 for foals born in the 1950s to 14.1 (North American racing only) for foals born from 1985-'94. Critics and apologists have cited everything from changes in training methods to the increased value of Thoroughbreds as the root cause of this vast decrease. For some, it seems that anything but soundness is to blame.

Still, the conclusion seems inescapable that the American Thoroughbred is less sound today than it was 50 years ago.

Obviously, there is still a smattering of foals born in 1985-'94 that are still racing, still out there gamely adding starts to the totals for their decade, gradually pushing up that 14.1-start average. Equally obvious is that most of them are geldings, a fact that leads to an interesting question, the answer to which might go a long way toward eliminating one factor often cited to explain the reduction in number of starts: residual value.

Geldings do not have any. Do not tell that to the thousands of horse lovers who gladly acquire Thoroughbred geldings every year as performance horses, riding horses, or simply to give them a good home. But to the Thoroughbred breeding industry, retired geldings have absolutely no commercial value. That is why such organizations as the Thoroughbred Retirement Foundation and California Equine Retirement Foundation exist.

Only such philanthropic foundations or the kindness of friends and strangers provide any value at all to the majority of geldings whose racing career is at an end.

Since geldings have no residual value, comparing the pattern of average number of starts for geldings over the last five decades to the pattern for all North American Thoroughbreds could give us an idea as to the role played by residual value in the reduction in average number of starts.

The table below compares percent starters, percent winners, average number of starts per foal, and average number of starts per starter for geldings and all foals regardless of gender born in the 1950s, '60s, '70s, '80s, and 1985-'94, the most recent ten-year period for which data are sufficiently complete. All data are restricted to North American racing, meaning that the results for all foals are somewhat different from the "Averages for the Breed" table that appears weekly in the stakes section of Thoroughbred Times because that includes worldwide starts. It should also be noted that Jockey Club Information Systems data as to which male horses were gelded is less complete in the 1950s, '60s, and '70s than it is now. The racing information on those horses recorded as geldings, however, is just as complete and, given the large numbers of horses involved, should not affect the results significantly.

So, does residual value have a significant effect on the number of starts Thoroughbreds make? Yes and no.

Declining numbers

Yes, geldings average many more starts than non-geldings. In the 1950s, geldings averaged 53.6 career starts recorded in the Jockey Club database, regardless of whether they ever made it to the starting gate. Exactly 90% of the geldings recorded were horses that made at least one start, and they averaged 59.7 career starts.

That is almost 20 career starts more per foal than all Thoroughbreds regardless of gender. That's 57.6% more career starts per horse with no residual value compared to all Thoroughbreds born in the same decade, regardless of residual value.

Over the next 3 1/2 decades, the average number of career starts for geldings recorded in the Jockey Club database has followed a pattern very, very similar to the pattern previously published for all Thoroughbreds foaled in that period-a precipitous decline. Compared to horses born in the 1950s, average number of starts in North America for Thoroughbreds born between 1985-'94 has declined 58.5% to 14.1. Over that same time period, average number of starts in North America for geldings has declined 47.8% to 28.

In the 1970s, '80s, and '90s, more North American Thoroughbreds were exported for racing than in the 1950s and '60s. It may be fairer, therefore, to compare figures only for those geldings and all foals that actually made it to the races in North America.

Those figures are equally discouraging about soundness. Average number of starts for geldings that started at least once has declined 48.4% since the 1950s, slightly more than for all geldings. Average number of starts for all North American-foaled Thoroughbreds that started at least once in North America has declined by only a slightly higher percentage, 49.8%, over the same time period. Obviously, most of the difference in percentage decline between all geldings and all foals is accounted for by the larger percentage of horses exported in recent decades.

Thus, it seems clear that lack of residual value does not exert anywhere near as strong an effect as does some other factor.

In other words, it is not the money. Or, at the very least, the data make it clear that the enormous increase in value of the average Thoroughbred since the 1950s cannot account for a substantial percentage of the huge decrease in number of starts per horse since that era.

Amounts contributed by several other factors, including innate soundness, liberalized medication rules, changed training methods, changes in animal husbandry, etc., cannot be determined from the data.

Equivocal interpretations

The data also offer at best equivocal support for those who believe that the current American Thoroughbred is just as sound as it ever was and that liberal medication rules enable horses to race longer and more often. If there is any segment of the Thoroughbred population that should benefit most in theory from medication, it is the thousands of geldings that populate the lower claiming ranks. In the 1950s, geldings averaged 57.6% more starts than the average Thoroughbred. All geldings foaled in 1985-'94 average 98.6% more lifetime starts than all foals. Among geldings and foals that started at least once in North America, however, the difference is only 43.3% more starts. Again, about half of the difference between all foals and all geldings is accounted for by horses exported.

But is the rest of that gap caused more by residual value, liberal medication, some other factor, or a combination of all of the above? Since geldings now make only a little over half as many starts as they once did, it is hard to argue that even the most liberal medication rules make them as sound as they once were. Most of the medications now in use did not exist in the 1950s and the cruder substances then available were much easier to detect.

Whatever the cause of the decrease in number of average starts by American Thoroughbreds, it seems logical that the percentage of that decrease attributable to commercial value cannot exceed something in the region of 15%.

How much of that decrease is attributable to a decrease in innate soundness is less clear.

Although most racetrack veterinarians and trainers are convinced that horses in their care can race longer and more often on therapeutic medication, the best data available do little to support that theory. In fact, the only published studies that have directly compared horses racing on specific therapeutic medications and those that do not have found that horses that do not race on medication have longer careers. Since about 95% of all horses now race on some type of medication, more definitive studies will be difficult to compile.

Racing secretaries, trainers, owners, and Thoroughbre