THE GENDER GAP: Elite Women Are Running Further Behind Stephen Seiler, Faculty of Health and Sport, Agder College, Kristiansand, Norway Steven Sailer, Writer and market researcher, Chicago, Illinois Sportscience News May-June 1997 http://www.sportsci.org/news/news9705/gengap.html

The world's fastest women are now getting slower, while the men keep getting faster. That's the outcome of our analysis of top running performances of the last 40 years. Reduced use of anabolic steroids may be the reason women runners are no longer becoming like men.

It's well known that top female runners have been catching up on their male counterparts. Some authors have even suggested that women might one day run as fast as men. For example, Ward and Whipp (1992) predicted that the men's and women's marathon world records would be equal by 1998 and that women record holders would catch up with men in shorter distances early in the 21st century. The American public agrees: a pre-Olympic poll of 1,000 adults last May found that 66 percent of Americans believe "the day is coming when top female athletes will beat top males at the highest competitive levels" (Tharp, 1996).

But hold on. Surely there is a hormonal gap between men and women that can't be closed? Their lack of testosterone means females have less muscle mass, higher body fat, smaller hearts, and less hemoglobin, even after correction for their smaller size. In top female athletes, these differences result in lower maximal oxygen consumption, less muscular strength, and less anaerobic capacity in comparison with the top males. The differences are still there after correction for body mass. Of course, if you gave women the missing hormones, it might be a different matter...

We began looking into this topic seriously after one of us (Seiler) casually informed the other (Sailer) that the steep improvement curve for women seemed to have flattened out. Unconvinced, Sailer suggested a closer look at the data. Together we analyzed the trends in elite running performance and found, to our surprise, that the women are now getting slower while the men continue to speed up. The reason? We believe that drug testing has reduced the use of anabolic steroids by both men and women. Drug testing is restoring the physiological gender gap.

To analyze running performance, we first considered previous analyses of world records (Hill, 1925; Ryder et al., 1976; Morton, 1983; Dyer, 1986; Ward and Whipp, 1992). We decided a more laboratory-like approach to the data was needed. In a laboratory, we would compare men and women under identical conditions using the same measurement instrument. With archival data collected since 1952, how is that possible? Well, we limited the analysis to running events where men and women perform under identical conditions. We used Olympic and world track and field championship finals to ensure that males and females were always running on the same track under the same environmental conditions (well, almost always). We used times from the top six finishers of each race, instead of the "sample size of one" that a world record represents. Since hand timing back in the 50s and 60s was a significant source of error in the sprints, we found the "true" times by using data from analysis of historical films of each race (Wallachinksy, 1996). To control for significant impact of wind on sprint performances, we assembled wind-velocity measurements from almost every race and corrected the raw times. (Thanks to Runners' World editor Amby Burfoot and Olympic historian Bill Mallon.) We corrected the sprints for altitude effects in the Munich and Mexico City Olympic Games. When political boycotts kept the best runners home, we threw out the data. And when we knew something unaccounted for was still impacting on the data, like the case of the 1960 Olympic 200-m final for women that was run in a sudden rainstorm, we also eliminated the data.

When all was said and done, we had analyzed the results of 182 championship finals (91 men's, 91 women's) from 12 Olympics and five IAAF world championships held between 1952 and 1996, 1091 data points in all. Here are data from one of the seven events we analyzed. We would say they are "representative" data, but the recent negative trend observed for females was actually more striking in some of the other sprint events.

Trends in 100-meter sprint times. Females' performances are represented by the top curve (red circles) and males' are represented in the bottom curve (blue squares). Each data point represents the mean value for the 1st through 6th place finishers' times after quantitative adjustment for hand timing, altitude, and wind velocity. Times from the 1952 men's final were omitted owing to a significant but unrecorded headwind. Times from both men's and women's 1980 Olympic finals were omitted because of the substantial impact of the political boycott on participation.

The correction methods filtered out a great deal of the fluctuation in performance in different events, and made small changes in recent years detectable. The performance curve above and all the others we generated were reduced to performance differences (the gender gap), expressed as a percentage of the male performance time. The gender gap in performance reached its nadir (minimum) in the 70s for the 800 m and in the 80s for the other events except the marathon.. The marathon is still showing a narrowing of the performance gap in the 90s, owing to slower performance by males (from an average time of 2:11.30 to 2:14.21), as female times are unchanged (2:30.02 vs 2:30.17). The absolute magnitude of the gender difference ranges from 9 to 13% across events and increases with distance. If the marathon is excluded, the mean performance gap for the six other running events has increased from 11% in the mid 80s to 12% in the mid 90s.

Although the data are not as clean, the progression of recent world records shows a similar divergence of females and males. Since 1989, world records for males have improved an average of 1.0%, whereas those for females have improved by 0.0% if we exclude a highly controversial track meet in Beijing and by 0.3% if we include that meet. Seven of the 10 men's world records standing in 1989 have been broken in the 90s. Men's world records set in the 80s had been bettered a total of 23 times by December 1996. In contrast, women's world records set during or before 1989 have been broken only once or three times, depending on whether the two controversial Chinese records are considered legitimate.

In short, the women are not still gaining on the men. The gap isn't even stable. The world's fastest female runners have become slower in the 90s! Is the widening of the gender gap in running "significant"? Statistically, the question is irrelevant. We are comparing two populations of world elite athletes, not samples. The increase in the gap is real.

Now for the obvious question: what has caused the turnaround? The measurement corrections we have applied would not favor males. In fact, effects of wind velocity have tended to favor male performances more than female performances, so our corrections have tended to reduce the gender difference. Factors that have improved performance in recent years, such as better track surfaces and improved aerodynamics of clothing, should benefit males and females equally. If measurement or environmental variables do not account for the deterioration of female performances in the last decade, social and cultural issues need to be considered. A number of such factors are presented in this table:

Overall continued improvements in health and nutrition, a relative increase in female elite sports participation, and increased professionalism all favor continued improvement in female performances, although not necessarily at the fast rate observed in the 60s and 70s. Only one significant social-cultural development could theoretically lead to deteriorating female performance at the highest levels. That development is improved drug testing. Random, unannounced drug testing was introduced in track and field in 1989, at least partly in response to the embarrassment of the Ben Johnson case in the 1988 Olympics. The single most powerful explanation for the growth of the gender gap in running is that the use of banned performance-enhancing drugs has declined as a result of this testing program, and that drug use, including anabolic steroid use, has previously benefited female performers more than males. Drug testing did not begin in 1989: drug tests were introduced as early as the 1964 Olympic games. However, early limits in test sensitivity, plus the fact that the tests were associated only with major events, blunted their effectiveness. It would be naive to think that drug use has come to an end as a result of more sensitive drug testing and random, unannounced testing methods, but the data presented here are evidence that a positive impact has been made.

Previous studies have explained the more rapid improvement in female running performances compared to men as a consequence of liberalization of social attitudes towards women in sport throughout the world, along with increases in participation and training quality at all levels of women's competition. This phenomenon was particularly evident from the 60s through the 80s. But if we inspect the running performances from the 70s and 80s more closely, the worldwide improvement was actually highly regionalized. Former East Germany, the Soviet Union, and the Former Communist Bloc dominated women's running in the 70s and 80s, taking nearly 70% of the final placement points in those two decades in Olympic and world championship races. That dominance ended abruptly in the 90s, coincident with the implementation of improved drug testing (and the collapse of the centralized systems of East Germany and the Soviet Union). Compared to the females, male performers from these countries never dominated running events. So, although their results also deteriorated slightly, the impact was less noticeable on the world stage.

The impact of masculinizing hormones on performance appears to have been far greater for women than men. The male performance data we have analyzed provides no evidence that improved drug testing since 1989 has even marginally impacted performance trends among the world's best male runners. World-class male performances have continued to improve at similar rates since 1989.

We conclude that the fall and rise of the gender performance gap in running is largely a consequence of the rise and fall in the illegal use of performance enhancing drugs from the 60s to the 90s. It is unlikely that drug use has been (or will ever be) eliminated, but it has been reduced.

References

Dyer, K. Catching up the men. (1984). New Scientist 1415(2), 25-26.

Hill, A.V. (1925). The physiological basis of running records. Lance 2, 481.486.

Morton, R.H. (1983). The supreme runner: what evidence now? Australian Journal of Sports Sciences, 3, 7-10.

Ryder, H.W., Carr, H.J., and Herget, P. (1976). Future performance in footracing.  Scientific American, 234, 108-119.

Tharp, M. (1996) Ready, Set, Go. Why we love our games. US News & World Report, July 22.

Wallechinsky, D. (1996). Sports Illustrated presents the complete book of the Summer Olympics. Boston: Little, Brown and Company

Whipp, B.J. and Ward, S.A. (1992). Will women soon outrun men? Nature., 355, 25.