Original Article

Partner wealth predicts self-reported orgasm frequency in a sample of Chinese women

Female orgasm in humans is a topic of continued debate and speculation but has relatively little solid empirical evidence (Dunn et al., 2005, Meston et al., 2004). Although there have been a multitude of definitions and physiologists, sexologists, doctors, and sociologists continue to debate the exact nature of the experience (Levin, 2004), there is general agreement that female orgasm is “a variable, transient peak sensation of intense pleasure, creating an altered state of consciousness, usually with an initiation accompanied by involuntary, rhythmic contractions of the pelvic striated circumvaginal musculature, often with concomitant uterine and anal contractions and myotonia that resolves the sexually induced vasocongestion (sometimes only partially), generally with an induction of well-being and contentment.” (Meston et al., 2004, p. 174).

The frequency of orgasm has been found to be an important component of sexual satisfaction, which in turn is a predictor of relationship satisfaction, for Chinese women (Parish et al., 2007). In American women, age and religiosity are negative predictors of orgasm frequency (Laumann et al., 1994, Meston et al., 2004), and the frequency of masturbatory orgasms but not orgasms with a partner increases with increasing education (Laumann et al., 1994). Twin evidence also suggests a moderate heritable component (Dunn et al., 2005). Beyond these, however, there is a relative paucity of evidence on the psychological and social factors influencing orgasmic function (Meston et al., 2004) since research has mainly focused on the physiological mechanisms involved (Levin and Wagner, 1985, Masters and Johnson, 1966).

Evolutionists have taken opposing positions on the function of female orgasm. On one hand, it has been seen as a functionless by-product of the ejaculatory response in males (Gould, 1987, Symons, 1979). An alternative view is that women's capacity for orgasm is an adaptation that serves to discriminate between males on the basis of their quality (Alcock, 1987, Smith, 1984, Thornhill et al., 1995), leading to either enhanced conception probability (Baker & Bellis, 1993) or selective emotional bonding with high-quality sires (Alexander, 1979, Fisher, 1992, Hrdy, 1996). While there has been considerable debate surrounding these hypotheses, there are only a few studies that set out to test the predictions of adaptive versus by-product hypotheses.

If female orgasm is adaptively designed for discriminating male quality, then it should be more frequent in females paired with high-quality males. There is evidence consistent with such an effect in Japanese macaques, where females display the orgasm-like clutching reaction more often when mating with high-ranking males (Troisi & Carosi, 1998). Two studies on humans have also found comparable effects. Thornhill et al. (1995) found that young women reported copulatory orgasm more frequently if their partners were more symmetrical. Shackelford et al. (2000) found that women paired with more attractive males were more likely to report orgasm at their last intercourse than women paired with less attractive males. These studies are interesting but rely on modest samples (n=86 and 388) of student volunteers, which means that the ranges of ages and male qualities may be quite limited. They also found effects only of symmetry and physical attractiveness as proxies of male quality. In a larger population cross-section, other qualities such as income and height may prove to be important.

In this study, we investigated for the first time in a large population cross-section the relationship between women's self-reported orgasm frequency and the characteristics of their partners. The data are from the Chinese Health and Family Life Survey, which was a large-scale investigation of sexual attitudes, behavior, and sexually transmitted diseases in a nationally representative sample of the Chinese population, carried out in 1999–2000 (Parish et al., 2003, Parish et al., 2007). For male quality, we investigated two parameters not only because they were available but also because previous population-level surveys have found them to be important determinants of male reproductive success. The first is income: Wealth is seen as desirable in potential partners by women all over the world (Buss, 1989), and studies in diverse populations have found that increasing wealth increases male marriage success (Borgerhoff Mulder, 1990, Fieder and Huber, 2007, Hopcroft, 2006, Nettle and Pollet, Pollet and Nettle, 2008). Thus, other things being equal, richer men are preferred to poorer ones as mates.

Our second independent variable is height: Several studies have found either a preference expressed by women for taller men or an actual increase in marriage success in taller-than-average men (Hensley, 1994, Mueller and Mazur, 2001, Nettle, 2002, Pawlowski et al., 2000) in Western populations. Little is known about male height and reproductive success in China. A recent study suggested that male height is not significantly positively related to offspring count, after controlling for potential confounds (Fielding et al., in press), yet height is generally positively related to health and socioeconomic status in China as it is elsewhere (Chen et al., 1990, Chen and Zhou, 2007, Samaras et al., 2003). Therefore, it is possible that height is a sought-after characteristic in a man for Chinese women as it is for women in Western populations.

If the adaptive view of female orgasm is correct, then we predict that women will report more frequent orgasms the richer their partners are and the taller their partners are. However, there is the danger that associations between partner characteristics and orgasm frequency might be due to confounding third variables. For example, the partners of richer men might be younger, be more educated, be physically healthier, or have higher levels of psychological well-being, or be more westernized, than partners of poorer men. In our analysis, therefore, we needed to control for a large number of potential confounding variables of this type.