tales from urban dilettantia

Icon

Wastage, Bleeders, and Murky Data in the Horse Racing Industry

Hey Australia, it’s Melbourne Cup Day.

It turns out that it’s actually quite hard to find well-researched information on issues of animal welfare in the racing industry.  All credit to organisations who lobby against cruelty in the industry, but their sites aren’t always the best of source of resources, and at times show a misunderstanding of the underlying statistics.

Given it’s Cup Day, I’ve put together an overview of some of the studies I’ve encountered during a quick skim of the literature. Bear in mind that I haven’t looked in detail into the authors, nor methodologies used in the studies, so cite with caution.  The two issues that immediately arose when I ran a search were ‘wastage’ (the commercial term for horses lost to the racing industry) and ‘bleeders’ (horses suffering from exercise-induced pulmonary hemorrhage or ‘EIPH’).

One of the most concerning aspects in my opinion is how just murky and under-scrutinised this whole industry appears to be in this respect – pinning down solid, credible data is no simple task, even where suspicions have been raised that the industry may be the horsey equivalent of a puppy mill.   For example, there’s very little information relating to the origin of horses sent to abattoirs.  This in in part due to an glaring absence of record keeping, the complication of abattoirs frequently procuring horses via auctions rather than directly from from racing stables, and the fact that some of the relevant data (where it exists) is considered to be commercial in confidence.

An estimated 80 per cent of those horses that actually end up on the racetrack suffer EIPH – these horses are known in the industry as ‘bleeders’.  (Hinchcliff, K.W., et al., Association between exercise-induced pulmonary hemorrhage and performance in Thoroughbred  racehorses. Journal of the American Veterinary Medical Association, 2005. 227: p. 768-774.)  This is quite an interesting one statistically, as horses can either bleed from the windpipe or in the deeper lung area, with some commentary noting the former applies to ‘around half’ of all racehorses, and the latter up to 90 per cent.  Wikipedia references studies stating the proportion of racehorses suffering EIPH at some point in their career falls between 40 per cent and 70 per cent.   I’m not clear on the number of non-racing horses who suffer from exercise-induced pulmonary hemorrhage, although some of the literature implies it is significantly lower.

It’s been estimated  “pregnancy in 1000 Thoroughbred Australian mares produces only 300 horses which will actually race”.   (Bourke JM (1995) Wastage in Thoroughbreds. In ‘Proceedings from the Annual Seminar, Equine Branch, NZVA’. Auckland pp. 107-119. Veterinary Continuing Education, Massey University)  Where do they go?  One of the more nteresting and recent studies in this area – greatly impaired as it was by lack of industry data – is detailed in a 2008 paper published in the Equine Veterinary Journal.  Here’s an extract of the relevant section:

An assessment was also made on the possibility of collecting further data within the abattoir setting. In this study data was collected over three collection dates from 340 horses processed at an Australian abattoir. This occurred between November 2007 and January 2008. The data showed that 59.8% of the horses had a dental age of  7 years with the remainder (40.2%) being > 7 years. Observations of the types of brands present indicated that 52.9% of the horses processed had originated from the racing industry with 40.0% of the sample group carrying a Thoroughbred brand and 12.9% carrying a Standardbred brand. The remainder of the group (47.1%) had no visible brand.

Wastage or horse loss (Jeffcott, 1990; Bailey, 1998) occurs at all stages of the horse’s life, including prior to racing, and it is estimated that pregnancy in 1000 Thoroughbred Australian mares produces only 300 horses which will actually race (Bourke, 1995). Similar pre-racing wastage has been found in Standardbred horses (trotters and pacers). A survey conducted on the 1990 crop of Western Australian Standardbred foals (Dyer, 1998) reported that 29% of foals were unregistered while approximately 26% were registered but never raced. Of the unregistered foals, 25% died or were destroyed and in 13% of cases, the cause of death was deliberate destruction. Of the registered, unraced horses 15% died and deliberate destruction was the cause of death in 12% of cases.

Bourke (1995) has also estimated that approximately 33% of the Thoroughbred population of Victoria may be lost to wastage each year however, these wastage figures include all areas in which horses are lost to the racing industry (e.g. reproductive failure, death of foals, various training and racing injuries and those relinquished for slaughter: Bailey, 1998). Interestingly, a more recent survey of racehorse trainers in the 2002/2003 race year reported similar figures. Hayek et al. (2005) found that the total wastage rate for horses in training or racing was 39% for Thoroughbreds and 38% for Standardbreds. Of the 39% of Thoroughbreds which left a racing stable only 6% were reported to have been sent to a knackery while 17% of Standardbred horses were reported to have been sent to the same destination. However, as the authors noted these figures do not include horses which were sent to a slaughter plant via a more indirect route, that is being sent to auction and purchased by an agent buying horses for slaughter, so the exact number of Thoroughbreds and Standardbreds in the study group which were ultimately slaughtered remains unknown.”

Doughty, A., Cross, N., Robbins, A. and Phillips, C.J.C. 2009. The origin, dentition and foot condition of slaughtered horses in Australia. Equine Veterinary Journal 41, 808-811.

Additionally, some of the literature suggests that horses who are unsuccessful on the racecourse may transition into the more harmful sport jump racing – a spectacle banned in New South Wales, and recommended to be phased out elsewhere by an Australian Parliamentary Inquiry.   Clearly, in addition to wastage and health issues, not to mention the subjects of gambling and whipping, there’s also a whole discussion to be had about the ethics of meat production versus the breeding of animals for an entertainment industry and so forth.  But, given they’re currently running  a Race That Stops A Nation, that’s one for another day.

Women of Numbers, Unite

Note (01 May 2012): I may have strayed from my intention in writing this one, as I fear it has been misinterpreted in some quarters.  I know many, many women who are good data analysts, and great data analysts.  I’ve read many wonderful articles containing great quantitative research.  However, the the best of my knowledge there is still a black hole when it comes to women talking about data as a feminist issue.  Datafeminists, to coin an awkward term.  Let’s keep talking.

I’m a researcher. I am passionate about research. And yet I hated every moment spent researching this article.

Search for any combination of words including ‘feminist’ and ‘statistics’ and you’ll see what I mean. There’s no body of work around the importance and use of statistics and data in feminist writing; no discussion around sourcing and interrogating data, and effectively communicating the information derived. Similarly, it seems that feminist posts taking oft-cited statistics and subjecting them to robust analysis don’t exist, or are so overwhelmed by a torrent of vitriol that they are near impossible to find.

Vitriol, you say? The posts I came across while searching for material were dominated by comments like these:

“Feminists never tire from promoting their lies”
“Why Feminism’s Vital Statistics Are Always Wrong”
“You are better off ignoring feminist stats”
“Feminism is the main cause of divorce in America”
“Feminists falsify facts for effect”

There are traps here. To say ‘we should have tried harder’ and so to accept the vitriol and the shaming, and – abhorrently – to blame ourselves. To rage against the often raised (and often valid) point that women must unfailingly conform to a higher standard than men to prove themselves. I’m probably going to fall into a few of those traps, in spite of trying my best.  But regardless, I wanted to write this and release it into the wild, because poor data, lazy research are problems wherever they arise, and it genuinely matters to me that we give these things our best effort – particularly when they pertain to very issues that we care about the most.

So, the researching of this post was a falling into the void in popular feminist writing that lurks in the place of well-referenced, well-researched, statistically sound numbers. A void where I would hope to see women with a passion for statistics vigorously promoting and debating the use of quantitative data. Encountering instead, unreferenced statistics, unsourced numbers, sweeping conclusions based only on anecdotal evidence. I’ve worked as a financial analyst, and now as an economist. I aspire to be the best rationalist I can be, however imperfect my achievement. And it grieves me to see such a deficiency, a great disconnect between two things I hold dear.

It’s not that the figures, the assertions, the conclusions are necessarily incorrect. But even if a number pulled from the ether without verification happens to be correct, this does not validate the process used to derive it. Erroneous – or perhaps worse – fundamentally unverifiable numbers propogate without scrutiny. Consider a number of specific cases. (I apologise in advance for cherry-picking and do note that these too are, ironically, anecdotal. However, given the shortage of self-critique and self-correction in feminist analysis, today we will settle for cautionary tales.)

1. Joan Brumberg, historian and former director of women’s studies at Cornell University wrote in Fasting Girls: The Emergence of Anorexia Nervosa as a Modern Disease that there were 150,000 to 200,000 fatalities from anorexia nervosa in any given year. Brumberg was misquoting the American Anorexia and Bulimia Association which had stated that there were 150,000 to 200,000 sufferers of of anorexia nervosa in the United States in any given year.

This error might have easily been identified by checking with the National Center for Health Statistics, which gave a figure of 70 deaths from anorexia in 1990. However, widely read authors including Naomi Wolf in The Beauty Myth and Gloria Steinam in Revolution From Within uncritically cited Brumberg’s figure without seeking out the primary source. (Both authors issued a correction once the error was highlighted.)

Even when writer Christina Hoff Sommers pointed out the mistake, she herself made the error of uncritically taking the Centre for Heath Statistics figure, stating that the actual number of deaths from anorexia was “less than 100 deaths per year.” In not considering the sources of data used by the the National Center for Health Statistics (which happened to be death certificates) she failed to consider heart failure, suicide or other causes of death arising as a consequence of anorexia. In contrast, the [peer reviewed] study, The Course of Eating Disorders (Herzog et al, eds.) indicated that the long-term fatality rate might be closer to 15%. Recognising the mistakes of others does not make one immune to making one’s own, and as Sommers herself said, “Where were the fact checkers, the editors, the skeptical journalists?” And, to give credit where it is due, Sommers has been one of our more vocal watchdogs when it comes to accuracy and factual reporting.

2. The March of Dimes Foundation, a United States non-profit established to work for the health of mothers and babies provides another example. In November 1992, Deborah Louis (then president of the National Women’s Studies Association) posted a message to the Women’s Studies Electronic Board citing the March of Dimes Foundation, stating that, “according to [the] last March of Dimes report, domestic violence (vs. pregnant women) is now responsible for more birth defects than all other causes combined.” Peculiarly, the March of Dimes Foundation did not publish a report on this topic, and was not aware of any research supporting the statement. Indeed, Maureen Corry, director of the March’s Education and Health Promotion Program, said “We have never seen this research before.”

This did not prevent Patricia Ireland, then president of the National Organisation for Women, saying that “battery of pregnant women is the number one cause of birth defects in this country” on the Charlie Rose program in February 1993.

The misinformation then propogated though The Boston Globe, the Dallas Morning News and Time magazine before the error was traced to the founder of a domestic violence advocacy project, Sarah Buel of Harvard Law School. Buel had misunderstood a statement made by Caroline Whitehead, a maternal nurse and child-care specialist in North Carolina, who cited a March of Dimes study indicating that more women are screened for birth defects than are screened for domestic battery. Whitehead had made no comment on the connection between battery and birth defects.

3. In January in 1993 at a news conference held by a coalition of women’s groups, reporters were told that Super Bowl Sunday is “the biggest day of the year for violence against women.”  The reporters were futher told that 40% more women would experience domestic battery on that day. (More, one might ask, than on what other day?) Sheila Kuehl (California Women’s Law Center) had used a study conducted at Virginia’s Old Dominion University three years before. Again, the statistic propogated through the media, with Rober Lipsyte of the New York Times referring to the “Abuse Bowl.”

The following day, psychologist and author of The Battered Woman Lenore Walker claimed on Good Morning America that she had compiled a ten-year report that showed the sharp spike in violent incidents against women on Super Bowl Sundays. And the day after that, reporter Lynda Gorov reported in the Boston Globe that women’s hotlines and shelters were “flooded with more calls from victims [on Super Bowl Sunday] than on any other day of the year,” citing “one study of women’s shelters out West” that “showed a 40 per cent climb in calls, a pattern advocates said is repeated nationwide, including Massachusetts.”

When writer Ken Ringle from the Washington Post called Janet Katz, professor of sociology and criminal justice at Old Dominion and co-author of the study originally cited by Kuehl at the news conference, Katz said “That’s not what we found at all,” and stated that an increase in emergency-room admissions “was not associated with the occurrence of football games in general.”

When Lenore Walker was asked to provide details of the findings from her ‘ten-year study’ she declined to share them, saying “We don’t use them for public consumption, we used them to guide us in advocacy projects.”

4. Since the mid-1980’s statements have have proliferated to the effect that women represent one half of the world’s population and a third of its labour force, are responsible for two-thirds of all working hours, receive a tenth of world income and own less than 1% of all property.

The numbers appeared in 1984 in Robin Morgan’s introduction to a book called Sisterhood Is Global: The International Women’s Movement Anthology. I remember seeing them in pamphlets and on posters at university, some fifteen years later. The oldest known source for them is in an editor’s introduction to an issue of the journal Women at Work, published by the International Labour Organisation in 1978, which stated:

“A world profile on women, using selected economic and social indicators, reveals that women constitute one half of the world population and one third of the official labour force; perform nearly two-thirds of work hours; but according to some estimates receive only one-tenth of the world income and possess less than one-hundredth of world property.”

Unsourced. No explanation of the ‘selected’ indicators. No elaboration on where ‘some estimates’ might have come from, or what these might be.

In 2007, author Krishna Ahooja-Patel, the editor responsible for that statement back in 1978, published a book called Development Has A Woman’s Face: Insights from Within the U.N. where she mentions that the formula was her own, and that it was “based on some available global data and others derived by use of fragmentary indicators at the time, in the late 1970s.”

The assumptions underlying Ahooja-Patel’s numbers include a guess that women constituted 33% of the world’s formal workforce and data from ‘several countries’ (unspecified) that they earned 10% to 30% less than men. From this, she took the higher end of the range from the earnings data, rather than a midpoint, and calculated that a third of the world’s total income was earned by women.

Further, Ahooja-Patel’s only explanation of the assertion that women own less than one hundredth of the world’s property is that “if the average wage of women is so low, it can be assumed that they do not normally have any surplus to invest in reproducible or non-reproducible assets.” She cites “various UN statistics” as her source.

For more than a quarter of a century, these numbers have filtered down through publications, women’s groups, the media, the internet and more. Often, the primary source is never stated, giving a misleading impression as to the date, time and context in which they were originally provided. They have been endlessly repeated wherever the issues of women, money, work and property are raised. And yet in their unreliability and unverifiability, they do no justice to feminism’s most critical concerns.

These are tales in isolation, demonstrating the manner in which bad information can indiscriminately spread. Far worse, is how little we care; where are our wonderful, fierce women arguing in favour of excellence in research and analysis? Where are those well-known women who have played key parts in the tales above, warning and educating us by virtue of the lessons they’ve learned? Where are the feminist bloggers, clamouring for an end to apathy and lazy journalism?  They may be out there, but we do not help their voices ring loud enough for me to find them in the world.

We can do better than this. So much better. I know women who are ethicists, financiers, lawyers, economists, actuaries, librarians, curators, researchers, doctors, biologists, accountants, architects, engineers, chemists, anthropologists, writers, geologists, journalists, linguists, computer scientists, pathologists, mathematicians, political scientists and more. Intelligent women who know better than to take a number at face value, or to state a conclusion without credible support. Intelligent women who value quality and who wholeheartedly support a culture of honest analytical contribution and critique.

Sometimes, we are story-tellers. Anecdotes have a valuable role in sharing a message, in communicating a large picture to a small audience. But we are not only story-tellers. We are astoundingly well-educated, connected human beings, and that in itself is a great privilege – the children of a providential intersection of race, class, geography and more.

Do better, loudly and visibly. Because we are astoundingly clever and astoundingly well-educated, and there is no honour in doing less than the best we can.

Dream(team)ing of Standard Deviation

In the spirit of giving you fair warning, if you’re not into Australian Rules Football or into data analysis, move along before you taint your eyes with the horrible mash-up of the two that follows.

Now, fair warning given, anyone who has had the pleasure of me herding them into an inescapable corner and ranting at them about standard deviation will know that I enjoy playing AFL Dream Team during the football season.    There’s nothing quite like hanging over the barrier at a game to yell ‘Oi, ya lazy #^%#!  Kick it, don’t handball it!” at one’s star recruit.  (Particularly if you’re also yelling ‘TACKLE HIM!!1!’ at your other star recruit who is on the opposing team.)  But most of all, I enjoy it because it’s fundamentally a game of statistics, and there are few things I love so hard as I love stats.

And so, I have a bit of a summer project going on this year.  The thing with Dream Team is that there are a bunch of players that everyone will have because they’re obviously going to (a) rise in value or (b) be consistent.  These players can be picked out quite readily by skimming the media or the plethora of Dream Team blogs and other resources that have come into being over the past few years.  The two things that differentiate a great Dream Team player from a middle of the road one are trading strategy and picking up relatively cheap players who unexpectedly come good.

The trading strategy is something I messed up a little this season just gone and will be working on, but my off-season project is all about the latter – trying to determine whether there are any early indicators of players who are about to have a good season.  As a first step, I’ve gone through a bunch of data I’ve managed to scrape from the web and hacked together a bit of an Excel model to help me pick out a pool of players to study.  (It turns out – not unexpectedly – that there are a lot of players who have a respectable second season after a low-averaging start as a rookie, but very few players who exhibit a dramatic jump in form from middle-of-the-road to Dream Team gun in years two to five.  In fact, far less than one would believe, given all the blog and forum chatter around the elusive ‘breakout year’).

Having identified these players, I’m going to look in more detail at their averages, games played, consistency and so forth in the year immediately preceding their ‘breakout year’ to see whether they share any common characteristics not observable in non-breakout players.  As a sideline, I’m also going to look at the second-year players who have demonstrated a significant improvement from their rookie form, although I think the reasons for this (and the likely players) tend to be a bit more obvious to begin with.  Here’s a screencap of the work-in-process with a bit more detail around the proposed  methodology:

Key Objectives - Fresh meat

Yes, this is truly what I do for fun on my lunch-break.  I reckon it beats shopping for shoes by a factor of about eleventy million.