Today the Employment Non-Discrimination Act (ENDA) was reintroduced to Congress for consideration. The legislation would add sexual orientation and gender identity to the list of ways to try to get your supervisor fired categories in which hiring and workplace discrimination is forbidden. The law provides exemptions for religious organizations and businesses with fewer than 15 employees.
The legislation has been introduced to Congress regularly since the 1990s (transgender protection is a relatively recent addition) but hasn't passed even under a Democrat-controlled Congress, even with some modest bipartisan support. With the speed of "evolution" on gay marriage recognition by politicians in 2013, it may get its moment soon.
The Wikipedia page for the legislation is amusing in that libertarians are the only people listed as potential objectors to ENDA who aren't conservative Christians. A 2010 essay from David E. Bernstein published at the Cato Institute (which somebody linked to the Wikipedia entry) defends libertarians against accusations of racism or homophobia (and other bias claims) by pointing out exactly how much power these laws give the government over the citizenry:
[S]upporters of antidiscrimination laws typically focus on laws banning racial discrimination. They do so because opposition to race discrimination has great historical and emotional resonance in a nation that had institutionalized racial oppression, including chattel slavery, for hundreds of years. However, federal antidiscrimination laws also apply to discrimination based on religion, sex, age, disability (including one's status as a recovering drug or alcohol addict), pregnancy, marital status, veteran status, and even military recruiters. State and local antidiscrimination laws cover everything from sexual orientation to political ideology to weight to appearance to membership in a motorcycle gang.
The proliferation of antidiscrimination laws explains why libertarians are loath to concede the principle that the government may ban private sector discrimination. There is no natural limit to the scope of antidiscrimination laws, because the concept of antidiscrimination is almost infinitely malleable. Almost any economic behavior, and much other behavior, can be defined as discrimination. Is a school admitting students based on SAT scores? That is discrimination against individuals (or groups) who don't do well on standardized tests! Is a store charging more for an item than some people can afford? That is discrimination against the poor! Is an employer hiring only the best qualified candidates? That is discrimination against everyone else!
The obvious retort is that antidiscrimination laws should be limited to "real" discrimination. But there is no consensus as to what constitutes "real" discrimination, nor, not surprisingly, does there appear to be any principled definition that legislatures have followed.
One can, for example, define discrimination as treating the alike unequally, but antidiscrimination law does not always follow this definition. Federal antidiscrimination law, for example, requires employers not simply to treat disabled and non-disabled alike, but to make costly "reasonable accommodations" for the disabled. Employers have the same legal obligation to their religious employees.
It's also difficult to determine how much discrimination against gay and transgendered employees is an actual thing anymore. Individual incidents will occasionally garner media coverage but calculations of actual workplace discrimination seems harder to come by. The Williams Institute at UCLA looked at states that have their own nondiscrimination policies protecting gay employees in 2011 and calculated the rate of complaints, which ended up being fairly similar to those for racial or sex discrimination, about 4 complaints per 10,000 gay workers (in both the public and private sector). The ultimate disposition of the complaints, though, does not appear to be calculated, so there's no way of knowing how many of those discrimination cases even proved to be valid.
The Human Rights Campaign, a major lobbyist for ENDA, seems focused entirely on whether or not the protections exist. When making note of the lack of workplace protection (and by the way, they show that 88 percent of Fortune 500 companies now have their own voluntary antidiscrimination policies protecting gay workers), there's seemingly very little effort to determine actual discrimination levels. There's no differentiating between the legality of antigay discrimination versus the reality of whether workplaces are engaging in discrimination even in the absence of government involvement. And the actual cases getting the most publicity (there's one today, in fact) often revolve around religious organizations (especially the Catholic Church), which will be exempt from ENDA anyway.
Yes, there is a lengthy history of anti-gay employment discrimination in the United States that goes back decades. But workplaces have "evolved" far faster than the government has. Even if we were to set aside libertarian support for freedom of workplace association and distrust of expanding government regulation (and I'm not suggesting you do), it's still worth asking the question of whether ENDA would actually accomplish anything culture and business economics isn't managing on its own.