Now, it's worth noting that the 1990 "gap" that Pew is reporting isn't all that big to begin with. A 6 percent difference is statistically significant, sure, but it isn't a wide gap. It's also worth noting that marriage by age 30 is dropping overall since 1990. Pew has several reasons for this: "declining economic fortunes of young men without a college degree and their increasing tendency to cohabit with a partner rather than marry."
This seems reasonable. Marriage -- or rather -- weddings are getting more and more expensive. With an economic recession that has resulted in particularly high unemployment among the under-30 crowd, it's no surprise that couples who are making a serious commitment might put off blowing a lot of cash on a party for their partnership. This was reflected in Amanda Marcotte's analysis of the New York Times' eye-rolling piece on the Millennials and why it takes them so long to "grow up" (i.e. make enough money to live on one's own and get married). Although some studies suggest that co-habitation before marriage is correlated with a higher divorce rate, that fact seems to have done little to deter people from taking the step. It's also important to note other studies have found a historical increase in co-habitation over time (PDF).
Still, I think at a fundamental level, this statistic about marriage is about larger social trends that have separated marriage and sex. Or at least, marriage and one's first sex partner. Much as conservatives huff and puff about how sex before marriage is the devil, even funding ineffective abstinence-until-marriage programs in middle and high schools throughout the country. But the more mainstream point of view is that linking that first sexual experience to marriage is not only uncommon, it's probably a bad idea. Marriage and relationship councilors often preach about sexual compatibility and how essential it is to a good marriage. Not knowing whether a partnership is sexually compatible before locking it into a legally binding agreement, it seems, is often perceived as a mistake.
I can anecdotally support Pew's theory. In my group of late-20s and early-30s friends, most are in serious relationships, but save the few that have recently gotten engaged, no one seems in a rush to the alter. Instead, nearly all are co-habitating. (Note that this is very different among my high school graduating class in the upper Midwest, which I suspect has a high pre-30 marriage rate. Some of these differences might be between urban and rural populations.) Still, this try-it-out trend among 20-somethings (and even 30-somethings and 40-somethings) is probably a good thing. Making commitments is great. But making commitments you're sure about because you've tried it out first is even better.