From the UK based The Marriage Foundation publication Divorce rates have halved for new brides. Why?:
Because it is almost entirely the reduction of wife-granted divorces concentrated into the early years of marriage that accounts for the overall 22% reduction in divorce rates since the 1993 peak, any explanation for this phenomenon has to account for wives being less prone to divorce. By far the most plausible explanation relates to wives perception of husbands.
In other words, husbands are doing better during the early years of marriage.
Why is this? Because commitment matters. He continues (emphasis mine):
The best current theory of commitment holds a plausible explanation for why this might be the case. Commitment theory proposes that men – but not women – who “decide” rather than “slide” through important relationship transitions are more dedicated and therefore have more stable relationships (Stanley et al, 2010).
The paper he cites to back up his assertion that men’s commitment matters in marriage and women’s doesn’t is Commitment: Functions, Formation, and the Securing of Romantic Attachment. Strangely I can’t find such a claim being made in that paper.
Still, the original Marriage Foundation paper is worth a look for the charts. The chart titled “Number of divorces (by party to whom granted)” on page 8 shows how initiation of divorce by men and women was roughly the same (with women initiating slightly more) until 1973. At that point women’s initiation of divorce skyrocketed and men’s initiation leveled off. Perhaps some of my readers from the UK can shed some light on what happened around 1973 to cause this change.
See also the chart titled “WIFE divorce rates, 2010 vs 1993” on page 9.