You'd probably want to control for wealth to get good data. Marriage duration is correlated with wealth, and wealthy people probably spend more on everything, weddings included.

[edit] nevermind, the the posted link claims the figures are controlled for several factors, including income and education, which both correlate with wealth.