Burnham et al (2003) found that men in current committed relationships (whether married or unmarried) had 21% lower saliva testosterone levels than single men.
Gray et al (2004) confirmed these results, and found that this effect was more pronounced in afternoon and evening samples of testosterone levels than in morning samples. However, it is not clear whether partnering causes lower T levels, or whether lower T levels increase the chance of partnering.
Van Anders and Watson (2006) found that single people with lower testosterone levels in Phase 1 of their study were more likely than high-testosterone subjects to be partnered at follow-up (6+ months later). Interestingly, van Anders and Watson also looked at testosterone levels in women, and took sexuality into account. They found that lower testosterone levels were associated with partnered status in heterosexual men, and non-heterosexual women, but not with heterosexual women or non-heterosexual men.
I've been lucky enough to be in one stable, committed relationship since before I began injecting testosterone. This makes me fairly rare among the transmen I know. Transition is necessarily a stressful, self-absorbed period in a person's life, and this tends to make maintaining a relationship difficult. The suddenness of the hormonal changes (in the case of ftms, the simultaneous onset of menopause and second puberty) can make the most trivial interactions difficult, as well. Therefore, I doubt the effect of hormone administration on the relationships of transsexuals is a good comparison model for hormone levels in the general population.