Can we all please stop claiming that “half of all marriages end in divorce”? Not only is this not supported by any real statistical analysis, but it is also a perfect example of how data can be skewed to give an unreasonable picture of the facts. I will be using absolutely no actual facts or figures to demonstrate my point. In fact, I will even cede the fact that there was one year (traditionally believed to be in the late 70’s) where there were half as many divorces as marriages, and even allow the extrapolation of that “fact” to apply to a longer period of time.
As an example, my parents have been married for over 40 years. And yet, those forty years make as much of an impact on the statistics of marriage as another member of my extended family who was married only once as well, but for just one week! (I did always love that woman, though) My parents could conceivably be married another forty years, and still the combined success rate for two members of the same family is only 50%. If we combine Elizabeth Taylor’s eight marriages and my one, we are looking at a joint rate just above 10%. Is the imbalance here beginning to show?
My point is that if we decide that it is important to determine a reliable statistic on the success rate of marriage, then the stats should be weighted. Even if someone gets married twice (but divorced only once), then the relative length of the two marriages should be taken into consideration. Someone who gets married at nineteen, gets divorced at twenty-one, remarries at twenty-six and then dies still married at eighty-one should not be given a 50% success rate. By basic math, half of his marriages failed. But the success of the one (55 years) should outweigh the “failure” of the other (2 years). A given person entering into a marriage does not have a 50% chance of getting divorced, even if we accept the unsupported popular statistics.
Unless, of course, this person works in Hollywood.
 The actual fact is that even if there were half as many divorces as marriages in a given year, many of those divorces were for marriages that began in other years and even decades, and is as likely to be a statistical anomaly as anything else. However, for the purposes of this post, I’ll allow the claimed 50% rate.