The LGBT community and its allies should go back to the ballot only when there is a reasonable chance to win.

Putting the issue of same-sex marriage back on the ballot before we have made progress solving the two problems noted above carries a high risk of defeat at the polls. While elections always entail risk, it makes sense to take the steps within our control to minimize it.

That is because defeat is costly. The dollar cost is roughly the same—probably $40–60 million—win or lose, each time we compete seriously. If we require three or four tries before we win, we are talking about spending $150–200 million. Why spend that money when $40–60 million will suffice if we’re willing to hold ourselves to a high standard and prepare seriously for the next election?

Yet the dollar cost is the least of it. Living through these ballot measures is brutal, particularly for LGBT people. Psychologist Glenda Russell documented the individual and collective cost in her book Voted Out: The Psychological Consequences of Anti-Gay Politics. Russell was one of the first to examine the psychological trauma instigated by ballot measures that essentially ask the populace at large to validate LGBT individuals as equal members of society. Examining Colorado's vote on Amendment 2, Russell argues that such votes unleash a barrage of psychologically painful discourse in which opponents to equality represent the LGBT community as aberrant, dangerous, inferior, and deservedly unequal. Debate of this character leads gay men and lesbians to internalize homophobic messages and harbor fears about future discrimination.

Sharon Rostosky expanded upon Russell's work and investigated the effect of 2006 anti-gay marriage ballot measures on the psychological well-being of LGBT people. Rostosky, along with Ellen Riggle, Sharon Horne, and Angela Miller, conducted a national longitudinal study of 1,500 LGBT individuals, nearly 600 of whom lived in states where a marriage equality measure was on the ballot. Comparing responses six months before the election with those weeks after the vote, and between states with and without a ballot measure, Rostosky's team found LGBT people suffered a higher level of psychological stress as a "direct result of the negative images and messages associated with the ballot campaign and the passage of the amendment." General anxiety disorder rates increased by 248%, and rates of depression and alcohol abuse rose as well.

The findings of Russell and Rostosky should very strongly motivate us to minimize the number of ballot measures our community has to go through.

Sometimes, of course, we have no choice. Often, as with Prop 8, our opponents choose whether and when an anti-LGBT measure will be on the ballot. Then, we have to fight no matter the cost because failing to stand up to a bully and failing to stand up for ourselves is the most dangerous and self-destructive choice of all.

But when, as now, the pro-LGBT side in California gets to decide when to seek repeal of Prop 8, it is prudent to think hard about whether and how we can put ourselves in a significantly stronger position to win. For both practical and ethical reasons, we have a responsibility to the LGBT community in particular and to our progressive allies in general to do the hard preparatory work before we instigate the next vote.